framework,version,device,op_name,kernel_source,gemm_dtype,m,n,k,latency
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,16384,19.87571038140191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,16384,41.20542738172743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,12288,33.85193549262153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,51200,61.358988444010414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,12288,15.089757283528646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,51200,137.1552734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,16384,84.63608805338542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,10240,32.59278361002604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,12288,61.248026529947914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,10240,12.781882392035591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,8192,20.91977267795139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,10240,51.14063856336805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,8192,10.35416497124566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,8192,43.88508097330729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,7168,18.10572984483507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,7168,9.389539082845053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,10240,36.64312065972222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,7168,35.764075385199654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,6144,17.570460001627605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,6144,7.818731519911025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,8192,29.684071858723957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,6144,30.712324354383682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,5120,12.494555155436197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,5120,6.74669689602322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,4096,10.340069240993923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,4096,5.3916443718804254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,4096,14.944193522135416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,3584,10.076838175455729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,51200,254.13441297743054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,5120,25.630406697591145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,7168,25.936331854926213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,3584,4.689911312527127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,3584,18.013153076171875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,3584,12.930120680067274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,4096,20.513159857855904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,3072,4.039006127251519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,3072,7.4224853515625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,2560,7.135393778483073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,2560,3.5262294345431857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,2560,8.158270094129774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,3072,13.756928337944878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,2560,8.337791442871094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,2048,5.33727306789822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,2048,3.031257841322157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,2048,6.221950107150608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,6144,22.325990464952255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,1536,5.0533396402994795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,1536,2.1340444352891708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,1536,4.649204254150391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,1536,4.608482784695096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,1024,3.2738827599419484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,1024,1.432750172085232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,768,2.5112862057156033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,1024,3.224111133151584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,768,2.4256470998128257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,768,2.6365636189778647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,768,1.1737741894192166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,512,1.6229670842488606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,512,1.0788115395439994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,512,1.9345874786376953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,256,1.260417726304796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,512,1.9852542877197266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,256,0.9517520268758138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,256,1.7472808625962999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,256,1.4413173463609483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,128,0.9180026584201388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,128,0.8362915780809191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,64,0.8166897561815051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,128,1.6808079613579645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,32,0.737055089738634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,64,1.7049866782294378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,32,2.183662202623155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,2048,6.123329586452908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,1024,3.2422559526231556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,128,1.0777351591322157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,3072,11.290848626030815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,65536,61.50145128038195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,51200,49.47986857096354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,16384,37.078162299262154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,65536,139.13663736979166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,51200,108.89222547743056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,16384,15.830518934461805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,16384,67.03457980685764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,16384,45.96221245659722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,12288,31.642754448784725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,12288,11.907901340060763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,10240,26.742302788628475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,12288,35.72648790147569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,12288,48.41815863715278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,10240,10.349200778537327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,8192,17.89831373426649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,10240,40.06043497721354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,10240,29.6075676812066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,8192,8.049732632107204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,51200,145.8057861328125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,7168,14.098871019151476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,8192,34.207665337456596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,8192,23.292190551757812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,51200,198.87721082899304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,7168,7.332322014702691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,7168,20.514624701605904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,65536,188.6012505425347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,6144,13.297838846842447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,6144,6.330878363715278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,7168,27.972925821940105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,5120,5.286092546251085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,4096,9.945203145345053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,6144,23.948862711588543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,5120,11.886049058702257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,5120,14.686118231879341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,4096,4.42409430609809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,5120,20.07732476128472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,3584,8.795992533365885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,4096,16.13164774576823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,4096,11.702710469563803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,3584,3.760606977674696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,65536,279.46522352430554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,3584,10.566459655761719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,3072,3.2465262942843967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,3072,7.579585605197483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,3584,14.064658270941841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,2560,2.7168212466769748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,3072,8.784378899468317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,2560,5.96707280476888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,3072,10.776775783962675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,2560,6.8251465691460504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,2048,2.0508755577935114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,2048,4.87375979953342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,2560,6.630118052164714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,1536,1.4438844256930883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,2048,4.9888691372341585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,2048,4.777584075927734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,1536,3.3819457160102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,1024,2.723212348090278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,1024,1.0931786431206596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,1536,3.6574427286783853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,1536,3.5638584560818143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,768,1.8283538818359375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,1024,2.5944027370876737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,1024,2.6732853783501516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,768,0.9036515553792318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,768,2.015067630343967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,512,1.3519066704644098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,768,2.0965493520100913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,512,0.7818426556057401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,256,0.8630133204989963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,512,1.5349884033203125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,512,1.5692301856146917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,256,0.6991724438137479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,256,1.3721840116712782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,256,1.16450129614936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,128,0.6664622094896104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,64,0.5470115343729655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,128,1.335627555847168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,32,0.5650151040818956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,128,0.7402097913953992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,64,1.3329102198282878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,32,1.7102808422512477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,6144,17.756981743706596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,65536,20.64054701063368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,51200,34.31987169053819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,51200,16.14083014594184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,51200,39.120459662543404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,65536,44.33136325412326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,128,0.8467840088738335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,16384,11.42429945203993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,16384,5.28459718492296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,12288,7.739595201280381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,16384,12.55035654703776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,12288,4.0072275797526045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,12288,9.422176784939236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,16384,21.341373019748264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,12288,15.652552286783854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,10240,3.1347137027316623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,10240,8.253125508626303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,10240,7.804312811957465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,51200,65.62624782986111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,8192,5.900525410970052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,8192,2.2166809505886502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,10240,12.960860358344185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,65536,50.42493693033854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,7168,5.069081624348958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,8192,6.287338680691189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,7168,1.8911181555853949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,8192,11.065816243489584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,6144,1.5310640335083008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,7168,5.460903167724609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,6144,4.597328186035156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,5120,1.1914133495754666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,6144,4.712332407633464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,7168,9.037794325086805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,5120,3.7355219523111978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,6144,6.5421252780490455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,4096,1.0535590913560655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,5120,3.73618655734592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,5120,4.300133175320096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,4096,3.1669705708821616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,4096,2.887255138821072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,4096,3.033715565999349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,3584,0.8519582218594022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,3584,2.6021751827663846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,3584,2.4809661441379123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,3584,2.5368614196777344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,3072,0.7183137999640571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,3072,2.027684529622396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,3072,2.2303519778781467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,2560,0.6896497938368055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,2560,1.9424834781222875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,2560,1.7220737669203017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,3072,2.191657808091905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,2560,1.8172151777479384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,2048,0.6407199965582954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,2048,1.6756667031182184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,1536,1.153769810994466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,2048,1.4378791385226781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,1536,0.4451155662536621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,2048,1.4447298049926758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,1536,1.0781510670979817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,1024,0.34494490093655056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,1024,0.8109448750813802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,1536,1.104928864373101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,768,0.5516471332973903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,1024,0.7787120077345105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,512,0.4046791129642063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,1024,0.8321386443244086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,256,0.2918399969736735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,512,0.39995643827650285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,768,0.6193928718566895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,768,0.2947102122836643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,512,0.7956870926751031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,128,0.2189342180887858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,512,0.5111893547905816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,256,0.47210221820407444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,128,0.2438275549146864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,64,0.17673422230614555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,256,0.4229653411441379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,128,0.4281075530582004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,32,0.1828604406780667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,64,0.4319422245025635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,32,0.4782569143507216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,256,0.23895822630988228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,768,0.6173297564188639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,65536,89.84942626953125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,65536,15.60704379611545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,51200,28.04892137315538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,51200,12.312513563368055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,65536,33.11241658528646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,51200,29.890567355685764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,16384,8.743298000759548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,16384,3.9337662590874563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,65536,38.22356160481771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,16384,9.614654541015625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,12288,6.487993876139323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,12288,2.763530731201172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,16384,16.532284206814236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,12288,7.2027392917209205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,10240,5.747529771592881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,12288,11.812397427029081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,10240,1.9355263180202906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,128,0.28808532820807564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,51200,49.569420708550346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,8192,1.4580266740587022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,10240,5.958584679497613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,8192,4.416920132107205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,7168,3.80000008477105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,8192,4.740494198269315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,7168,1.2010889053344727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,10240,9.810105217827692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,8192,6.399579366048177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,7168,4.569702996148004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,7168,3.9416380988227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,6144,1.0016933017306857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,5120,0.9401537577311198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,6144,3.4705545637342663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,6144,3.5749664306640625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,6144,3.319511201646593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,5120,2.7703751458062063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,5120,3.142815907796224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,5120,2.7125297122531467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,4096,0.6958773401048449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,4096,2.4231209225124783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,4096,2.07223023308648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,3584,0.6842133204142252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,3584,2.063640806410048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,4096,2.198388417561849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,3584,1.7856080796983507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,3584,1.886794620090061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,3072,0.6158746613396539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,3072,1.5194649166531031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,2560,0.49355554580688477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,3072,1.9998577965630426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,2560,1.586771541171604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,3072,1.6595670912000868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,2560,1.264892472161187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,2560,1.3674613104926214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,2048,0.520159986284044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,2048,1.1847510867648654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,1536,0.8695199754503039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,2048,1.044223149617513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,1536,0.8065599865383573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,2048,1.0729431576199002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,1536,0.38276534610324436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,1024,0.7659564548068576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,1536,0.795888000064426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,1024,0.2613137827979194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,768,0.43858755959404844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,1024,0.5959591335720485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,1024,0.5782426728142632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,768,0.23939288987053764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,512,0.38203022215101456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,512,0.37966399722629124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,512,0.2086248927646213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,256,0.2899128860897488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,768,0.47268978754679364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,768,0.733032915327284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,128,0.1596471071243286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,512,0.37965067227681476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,256,0.26317956712510854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,256,0.33789422776963973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,64,0.19495911068386504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,256,0.2851777871449788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,128,0.3284906546274821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,128,0.19013599554697672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,32,0.20245777236090767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,64,0.3216666645473904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,32,0.3501608901553684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,65536,67.01825629340279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,65536,13.316112094455294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,51200,22.753134833441838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,51200,10.564283582899305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,65536,29.03192138671875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,16384,7.4093822903103295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,51200,25.136162651909725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,16384,3.294514762030708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,65536,32.55992635091146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,16384,8.164290534125433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,12288,5.857942369249131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,16384,13.596767001681856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,12288,1.963928010728624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,12288,6.084458669026692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,10240,1.6050293180677626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,51200,42.050811767578125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,10240,4.566793653700087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,12288,10.042818705240885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,10240,5.023263295491536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,8192,1.2544213400946724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,8192,3.867967817518446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,10240,7.2661327785915795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,8192,3.771540323893229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,8192,4.3411102294921875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,7168,1.0386533737182617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,7168,3.4195609622531467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,7168,3.332150141398112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,7168,3.1379884084065757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,6144,2.691450754801432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,6144,0.9888035456339518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,6144,2.7206177181667752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,6144,2.71433957417806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,5120,0.816673755645752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,128,0.22039377689361572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,4096,0.5920862091912164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,5120,2.4934630923800998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,5120,2.1657625834147134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,4096,1.924425760904948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,5120,2.253055148654514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,4096,1.6729350619845922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,4096,1.8007431030273438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,3584,0.6519110997517904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,3584,1.7125288645426433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,3072,0.5001600053575304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,3072,1.4439271291097004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,3072,1.2840995788574219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,3584,1.628519058227539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,2560,1.155112902323405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,2560,0.424743996726142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,3072,1.3581226136949327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,2048,0.9073706732855903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,3584,1.5447449154324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,2560,1.1090560489230685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,2048,0.362115568584866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,2560,1.1151946385701497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,1536,0.28572267956203884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,2048,0.9789181815253364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,1536,0.824760013156467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,1024,0.43291200531853574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,1536,0.6751839849683973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,1536,0.6708818011813693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,1024,0.3429768880208333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,768,0.36034488677978516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,1024,0.524621327718099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,768,0.3951084348890517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,512,0.27379022704230416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,1024,0.49506664276123047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,768,0.27473423216078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,256,0.16167199611663818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,512,0.3227573235829671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,512,0.25117243660820854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,768,0.40447200669182676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,2048,0.870556460486518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,256,0.2945200072394477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,128,0.13685688707563612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,512,0.322343111038208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,256,0.2851022349463569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,64,0.11256978246900771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,65536,57.044772677951386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,256,0.2505057652791341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,128,0.2667688793606228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,128,0.16294577386644152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,32,0.11847733126746283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,64,0.283930672539605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,32,0.28937334484524196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,65536,10.933926900227865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,51200,17.40459526909722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,51200,8.677586873372396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,65536,22.076010810004338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,16384,6.114316304524739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,51200,20.009245130750866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,16384,2.1967012617323136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,12288,4.3129556443956165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,16384,6.581602732340495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,12288,1.4879982206556532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,16384,10.989663018120659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,12288,6.834288703070746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,51200,32.83546278211806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,12288,4.871128929985894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,10240,3.9202168782552085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,10240,1.203703138563368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,10240,4.146803538004558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,8192,3.2594114939371743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,8192,0.9798950619167752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,10240,3.849767896864149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,8192,3.188627454969618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,8192,3.0029527876112194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,7168,0.8295511139763726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,7168,2.510045369466146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,7168,2.777000003390842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,7168,2.6098925272623696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,6144,2.274649726019965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,6144,0.725697782304552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,6144,2.272893269856771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,6144,2.167083528306749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,5120,1.932636472913954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,5120,0.6152248912387425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,5120,1.7749279869927301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,5120,1.828383975558811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,4096,0.4852275318569607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,4096,1.5880106820000544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,4096,1.5082551108466253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,4096,1.4189235899183485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,3584,1.5142666498819988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,3584,0.4325893455081516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,3584,1.2329769134521484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,3072,0.9934790929158529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,3072,0.38748976919386124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,3584,1.2768221961127388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,3072,1.055884467230903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,3072,1.0794426600138347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,2560,0.9089004728529188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,2560,0.34488622347513836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,2560,0.8861378033955892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,65536,26.328864203559025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,2560,0.8975244098239474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,2048,0.7352017826504178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,2048,0.28970932960510254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,2048,0.7648764186435275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,1536,0.2399804327223036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,1536,0.6100453270806206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,2048,0.7092115614149305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,1024,0.35274754630194766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,1536,0.5872880087958442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,1024,0.19468533992767334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,768,0.26837778091430664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,1024,0.4176817735036214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,1536,0.5692044364081489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,512,0.21274756060706246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,768,0.32790133688184947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,768,0.2908559905158149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,1024,0.4438942273457845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,512,0.15266222423977321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,256,0.21874933772616914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,65536,43.7764146592882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,512,0.26531023449367946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,256,0.24556888474358451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,128,0.10923022031784058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,256,0.19683467017279732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,64,0.08791377809312607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,128,0.14091822836134169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,128,0.21846044063568115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,32,0.09496000077989365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,128,0.15534488360087076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,64,0.2191013362672594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,512,0.2728808985816108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,32,0.2271182272169325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,768,0.3382417890760634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,256,0.2414728800455729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,128,0.1863786644405789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,65536,9.82873789469401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,51200,7.5011444091796875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,16384,5.017126295301649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,51200,16.92053392198351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,65536,18.37146165635851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,16384,1.8354843987358942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,51200,18.330108642578125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,16384,8.710560268825954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,16384,5.780883365207249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,12288,1.3315555784437392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,12288,4.024242824978299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,12288,4.105874803331163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,12288,5.028873019748263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,10240,3.414086235894097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,10240,1.0551360448201497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,10240,3.362885369194878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,10240,3.337263955010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,8192,0.8376248677571615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,8192,2.6381183200412326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,8192,2.5860907236735025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,51200,27.787105984157986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,8192,2.59443728129069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,7168,2.294326146443685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,7168,2.2126693725585938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,7168,0.7376009093390571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,6144,0.6245448854234483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,6144,1.9502506256103516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,6144,1.8905688391791449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,5120,1.5705440309312608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,6144,1.9878898196750219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,5120,0.5463591151767307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,7168,2.223923577202691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,5120,1.5735431247287328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,4096,0.4416400061713324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,4096,1.3493377897474501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,5120,1.5825653076171875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,4096,1.2087217966715496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,4096,1.2843537860446506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,3584,0.3959466616312663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,3584,1.095320913526747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,3072,0.9644035763210721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,3072,0.3498977820078532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,3584,1.1029795540703666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,3072,0.9631190829806857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,3584,1.0804595947265625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,2560,0.8696782324049207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,3072,0.9520329369439019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,2560,0.30718310674031574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,2560,0.7812799877590604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,2048,0.6342888938056098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,2048,0.26227111286587185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,2560,0.7747368812561035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,1536,0.45609331130981445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,1536,0.21308444605933297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,2048,0.6729404661390516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,2048,0.6326257917616103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,1024,0.3418568770090739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,1536,0.4991564220852322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,1024,0.17216622829437256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,1536,0.8018390867445203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,1024,0.3776657846238878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,768,0.16097599930233425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,512,0.19138399759928384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,1024,0.3647591008080377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,768,0.28984710905287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,512,0.13967288864983454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,256,0.1253875494003296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,768,0.2935217751397027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,512,0.23924265967475045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,128,0.0984613365597195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,256,0.1257208850648668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,512,0.24509244494967988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,256,0.20373866293165419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,64,0.07623022132449679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,256,0.17540978060828316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,768,0.259223116768731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,128,0.12761244508955213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,128,0.1902408864763048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,32,0.08083555433485243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,128,0.13515467113918728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,64,0.19264177481333414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,32,0.1990133391486274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,65536,23.378953721788193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,65536,39.23672824435764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,65536,8.664661831325954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,51200,12.202339172363281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,16384,4.200631035698785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,51200,6.482466803656684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,16384,1.5617013507419164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,65536,18.25805833604601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,16384,4.958704206678602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,16384,6.307334899902344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,12288,3.356440862019857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,12288,1.1219280030992296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,65536,20.026940239800346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,51200,20.996239556206596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,12288,3.605436537000868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,10240,3.074433856540256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,10240,0.8866115676032172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,12288,3.9157129923502603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,51200,15.998036702473959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,10240,3.034722646077474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,8192,0.8157182269626193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,10240,2.865799162122938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,8192,2.414808909098307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,8192,2.2867359585232205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,8192,2.2890906863742404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,7168,0.7978035608927408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,7168,1.9954026540120442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,7168,1.9099449581570094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,6144,0.5843911170959473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,7168,1.9681165483262804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,6144,1.817196528116862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,6144,1.7345973120795355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,5120,1.4019475513034398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,5120,0.46944623523288304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,6144,1.6798630820380316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,5120,1.3212862014770508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,4096,0.4054728878868951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,4096,1.1659182442559135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,4096,1.060516463385688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,5120,1.3704675038655598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,3584,0.90939818488227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,3584,0.42407645119561094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,4096,1.1608284844292533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,3584,0.9151715172661675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,3072,0.7493955294291178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,3072,0.32803556654188365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,3584,0.9635581970214844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,3072,0.8166346549987793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,2560,0.6561075316535102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,2560,0.307041777504815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,2048,0.5226773156060113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,3072,0.8421129120720757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,2048,0.23272532886928984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,2048,0.553874651590983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,2560,0.7105715539720324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,1536,0.460044436984592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,1536,0.18995911545223662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,1024,0.25597956445482045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,2048,0.6944373448689779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,1536,0.4366328981187608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,1024,0.18637600209977892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,768,0.2120915518866645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,1024,0.4924329121907552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,2560,0.667777803209093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,512,0.174255993631151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,768,0.1796426640616523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,1024,0.3192257881164551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,768,0.2534986601935493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,256,0.0991777777671814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,768,0.25867199897766113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,512,0.20277867052290174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,128,0.08374577760696411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,256,0.1769324408637153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,1536,0.44690222210354275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,256,0.11498578389485677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,512,0.20813422732883027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,64,0.06906844509972467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,256,0.15339022212558323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,32,0.0712346699502733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,128,0.11326932907104492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,128,0.1643093294567532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,64,0.1672382222281562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,128,0.1181653340657552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,32,0.1702124410205417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,512,0.13884088728162977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,65536,28.477518717447918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,51200,5.102291531032987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,65536,7.052583906385634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,16384,3.611243353949653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,51200,11.514251708984375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,16384,4.472969055175781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,16384,1.2866124047173393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,51200,13.587527804904513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,16384,4.169944763183594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,12288,2.680671903822157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,65536,15.812906901041666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,12288,0.8812417984008789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,12288,3.115901311238607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,10240,0.7483546468946668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,10240,2.770096037122938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,65536,17.227535671657986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,12288,2.9998736911349826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,10240,2.235955556233724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,10240,2.476295895046658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,51200,19.100723266601562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,8192,1.9435609181722004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,8192,1.7829297383626301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,8192,1.9491519927978516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,8192,0.6034657690260146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,7168,0.6000462108188206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,7168,1.711838192409939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,7168,1.5852933459811742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,6144,0.482477347056071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,6144,1.4758542378743489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,7168,1.6981422636244032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,5120,1.239180458916558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,6144,1.4328141742282445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,6144,1.4278586705525715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,5120,1.1906684239705403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,5120,0.5644959873623318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,4096,0.8049342367384168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,4096,0.33325245645311147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,5120,1.1738337410820854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,3584,0.7443226708306206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,3584,0.30144445101420086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,4096,0.8948053783840604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,4096,1.1669901741875541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,3584,0.8761777877807617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,3072,0.2658951017591688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,3072,0.7247893545362684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,3584,0.8073235617743598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,2560,0.5124328931172689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,3072,0.6903173128763834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,2560,0.26452888382805717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,3072,0.6966115633646647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,2048,0.4070231119791667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,2048,0.2016284465789795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,2048,0.5653075642055935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,1536,0.32403021388583714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,2560,0.7423573070102267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,1024,0.22448711925082734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,2048,0.48271910349527997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,1536,0.22386222415500215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,1536,0.37267110082838273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,768,0.19347555107540557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,1024,0.1381502283944024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,1024,0.284189330206977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,1536,0.3719973299238417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,512,0.13365333610110813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,768,0.12927999761369494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,2560,0.5819955401950413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,1024,0.27771467632717556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,256,0.08496889140870835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,512,0.11639911598629421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,512,0.17381510469648573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,768,0.27250221040513783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,128,0.06953955358929105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,256,0.14981244670020208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,512,0.183296004931132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,64,0.05643377701441447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,256,0.13135733869340685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,128,0.101492444674174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,128,0.14000977410210505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,128,0.10121599833170573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,64,0.13987733258141413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,32,0.14170755280388728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,65536,26.91455078125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,768,0.2138435575697157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,256,0.10219111045201619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,32,0.060439997249179415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,51200,3.8489333258734812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,65536,4.94523451063368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,16384,2.878064897325304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,16384,3.2118665907118054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,51200,8.946693420410156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,16384,1.0276560253567166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,65536,11.004083421495226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,12288,2.2266195085313583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,16384,3.389120101928711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,51200,10.875144110785591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,12288,0.7794906828138563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,12288,2.2706790500217013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,12288,2.408273696899414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,10240,0.6457991070217556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,10240,1.945540534125434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,10240,1.97742674085829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,8192,0.5060124397277832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,10240,2.011152903238932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,65536,14.57933383517795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,8192,1.5539635552300348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,8192,1.623059590657552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,7168,0.5513226721021864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,8192,1.6075973510742188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,7168,1.4714924494425456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,7168,1.344760047064887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,6144,1.1689315372043185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,7168,1.372131559583876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,6144,0.465361754099528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,6144,1.2600089179144967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,5120,0.4063119888305664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,6144,1.1942417356703017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,51200,17.29830593532986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,4096,0.27814311451382107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,5120,1.1365867190890842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,4096,0.6485448943244086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,5120,0.9434258143107096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,4096,0.7657377984788684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,3584,0.252378675672743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,4096,0.753876421186659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,3584,0.5840497546725804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,5120,0.9903963936699761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,3584,0.644960880279541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,3072,0.4945519765218099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,3072,0.2915182113647461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,3584,0.7293608983357748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,2560,0.4580400254991319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,3072,0.5802266862657335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,2560,0.2626986768510607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,2560,0.47300264570448136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,3072,0.592773331536187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,2048,0.3256302144792345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,2048,0.17308710681067574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,2560,0.5152302318149143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,1536,0.27195554309421116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,2048,0.6414097679985894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,1024,0.18251466751098633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,2048,0.4031004375881619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,1536,0.1458239952723185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,1536,0.43093334303961855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,768,0.1347928841908773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,1024,0.117576003074646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,1024,0.22306577364603677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,1536,0.3262293338775635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,512,0.11244800355699326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,768,0.11197155051761204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,1024,0.3007724550035265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,256,0.07194133599599202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,512,0.10065066814422607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,512,0.14518045054541692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,512,0.15049955579969618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,256,0.0907671120431688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,256,0.1228933334350586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,256,0.109351999229855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,128,0.1132062276204427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,64,0.046279999944898814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,32,0.04765599966049194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,768,0.18468088573879668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,64,0.11295110649532741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,768,0.18925510512457955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,32,0.11397866408030193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,128,0.055529776546690196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,128,0.089336891969045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,65536,20.549063788519966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,65536,4.31060791015625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,51200,3.5117225646972656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,51200,9.222672356499567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,65536,10.706195407443575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,16384,2.627304924858941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,16384,0.8948382271660699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,16384,2.8999671936035156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,65536,12.310679117838541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,51200,10.01748318142361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,16384,3.0031138526068797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,12288,0.739832878112793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,12288,1.9397883945041234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,12288,2.127861234876845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,10240,1.6305821736653645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,12288,2.222179624769423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,10240,0.6710489061143664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,10240,1.8276435004340277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,51200,15.48529052734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,8192,1.5095031526353624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,8192,0.4575626585218642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,65536,16.908484564887154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,8192,1.414232889811198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,10240,1.8550470140245225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,7168,0.42187823189629453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,7168,1.1884987089369032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,8192,1.4385582605997722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,6144,0.924957381354438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,6144,0.51200622982449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,7168,1.2359484566582573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,6144,1.000226656595866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,7168,1.221981366475423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,5120,0.7843093342251248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,5120,0.3015804555681017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,4096,0.5615146425035265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,6144,1.0682044559054904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,4096,0.25701867209540474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,5120,1.1903680165608723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,4096,0.6840195655822754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,3584,0.6011626455518935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,5120,0.8687404526604546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,3584,0.23143466313680014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,4096,0.7097919782002767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,3072,0.41942400402492946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,3072,0.20650222566392687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,2560,0.3559653229183621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,3584,0.6127493116590712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,2048,0.29020622041490346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,2560,0.22956265343560112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,3072,0.5347306463453505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,2560,0.4990533192952474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,2048,0.15692888365851507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,2560,0.4466400146484375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,1536,0.22582578659057617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,3584,0.5948373476664225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,1536,0.14187199539608425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,1024,0.16211288505130345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,1536,0.2751528951856825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,2048,0.3891022205352783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,768,0.1181466711892022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,1536,0.2872089015112983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,1024,0.10987377829021877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,1024,0.20812266402774385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,3072,0.7541013293796115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,512,0.0895528859562344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,768,0.10493333472145928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,1024,0.21294311682383218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,768,0.15879910522037083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,256,0.06429244412316217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,768,0.17151467005411783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,512,0.12819466325971815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,512,0.0941057801246643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,128,0.0484426650736067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,512,0.13549333148532444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,2048,0.39798845185173887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,256,0.08443111181259155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,256,0.10937422513961792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,64,0.04067822297414144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,128,0.0833137763871087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,32,0.042661332421832614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,128,0.10027377472983466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,256,0.09852977593739827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,128,0.07555111249287923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,64,0.09956977764765422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,32,0.10047022501627605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,65536,3.813347710503472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,51200,6.883989122178819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,51200,3.239301257663303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,16384,2.434334225124783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,65536,9.916580200195312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,16384,0.8235911263359917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,16384,2.6053875817192926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,51200,10.916058858235678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,51200,8.854274325900608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,65536,10.996361626519096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,12288,0.6117555300394694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,12288,1.6625804901123047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,16384,2.667538748847114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,10240,1.5067467159695094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,12288,1.911502202351888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,12288,1.9357386695014105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,10240,0.5101706716749403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,65536,13.276921590169271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,10240,1.5718364715576172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,8192,0.41231645478142637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,8192,1.1547199885050456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,7168,0.9896329243977865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,8192,1.2023004955715604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,10240,1.611787584092882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,7168,0.520343992445204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,8192,1.2514479955037434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,7168,1.0483919779459636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,6144,0.3594044314490424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,6144,0.9002728992038302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,7168,1.0988666746351454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,5120,0.637389341990153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,6144,0.9349493450588651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,5120,0.3542435434129503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,6144,0.9413661956787109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,5120,0.7522133191426595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,4096,0.49304713143242734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,4096,0.2997013462914361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,3584,0.4289235538906521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,5120,0.8129706912570529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,4096,0.6127120123969184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,3584,0.21004622512393525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,3584,0.5188177956475152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,3072,0.4112257692548964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,3072,0.18849155637953016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,2560,0.34254222446017796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,3072,0.45845868852403426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,2560,0.1836320029364692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,3072,0.4693102306789822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,2048,0.2597600089179145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,2560,0.5827440155877007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,1536,0.19746577739715576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,4096,0.6200213432312012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,2048,0.14955200089348689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,2048,0.33986488978068036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,1536,0.12299999925825332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,1024,0.13166933589511448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,2048,0.33409688207838273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,1536,0.2587120003170437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,3584,0.6888186666700574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,768,0.10259289211697048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,1536,0.2568568918440077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,1024,0.10160622331831191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,1024,0.18080266316731772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,512,0.08176889022191365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,768,0.09745066695743138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,1024,0.1924142175250583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,768,0.1400053368674384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,256,0.055267555846108325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,512,0.11394666300879584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,512,0.08744266960355972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,768,0.15297510888841417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,2560,0.4039528899722629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,128,0.03792266713248359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,256,0.07827555470996432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,512,0.12120266755421956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,64,0.03403822249836392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,128,0.07738488912582397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,32,0.03560622202025519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,256,0.08757599857118394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,128,0.08775199784172906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,64,0.08638399839401245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,32,0.08632266521453857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,256,0.09560977750354344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,65536,3.267854266696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,51200,5.560661315917969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,65536,8.289011637369791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,128,0.08406044377221002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,51200,2.665879143608941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,16384,1.8384693993462458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,16384,2.0734924740261502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,16384,0.7005955378214518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,51200,7.446726481119792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,12288,1.526892450120714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,16384,2.299220403035482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,12288,0.6391119956970215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,12288,1.593790266248915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,51200,10.381207784016928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,10240,1.1922008726331923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,12288,1.6476675669352214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,10240,0.4596586757236057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,10240,1.3061751259697807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,10240,1.4128612942165797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,8192,1.087135950724284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,8192,0.3588835663265652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,8192,1.0616436004638672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,8192,1.1206275092230904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,7168,0.3150160047743055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,7168,0.7831795480516223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,6144,0.6431999736362034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,7168,0.975936041937934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,7168,1.1605555216471355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,6144,0.3091964456770155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,6144,0.7774675687154134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,65536,12.0912594265408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,5120,0.5259448687235514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,5120,0.2383208804660373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,6144,0.92940796746148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,4096,0.4237831168704563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,4096,0.20070932971106636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,3584,0.39522843890719944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,5120,0.6798897849188911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,4096,0.5978959931267632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,4096,0.53874757554796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,3584,0.45695824093288845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,3584,0.18185155921512178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,3072,0.3386080000135634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,3072,0.16423022747039795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,2560,0.2824808756510417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,5120,0.6459617614746094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,3072,0.4138195514678955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,3584,0.5465946727328831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,2560,0.14414578013949925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,2048,0.22089866797129312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,3072,0.4192639986673991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,2560,0.3279120127360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,1536,0.166756444507175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,2048,0.14099110497368708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,2048,0.2752142217424181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,2560,0.42990756034851074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,1024,0.11439288987053765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,1536,0.1126284466849433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,1536,0.20960533618927002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,2048,0.2931911150614421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,768,0.09060266945097183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,1024,0.09433955616421169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,1024,0.15911555290222168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,1536,0.23034932878282335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,512,0.07100355625152588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,768,0.090756442811754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,1024,0.17029421859317354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,256,0.04517689016130236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,512,0.08094666401545207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,512,0.0983555581834581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,768,0.13536889023251003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,128,0.03654311100641886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,256,0.07264888948864408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,512,0.1070515579647488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,256,0.08197777801089816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,64,0.03166133496496413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,128,0.07107555866241455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,256,0.07643110884560479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,32,0.03362577822473314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,128,0.05862666500939263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,64,0.07292711072497897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,32,0.07259733147091336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,768,0.12199200524224176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,128,0.07452178001403809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,65536,2.7187493642171225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,51200,4.191807217068143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,65536,5.950005425347222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,51200,2.179138607449002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,65536,9.639867146809896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,16384,1.4363760418362086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,16384,0.6480506791008843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,65536,8.265703837076822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,51200,7.8003896077473955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,12288,1.0498586230807834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,16384,2.002725389268663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,51200,6.327530754937066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,12288,0.5142133500840929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,12288,1.3452897601657445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,10240,0.928383085462782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,65536,10.780523512098524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,12288,1.4180755615234375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,16384,1.824975119696723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,10240,0.5191813574896919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,10240,1.12337154812283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,8192,0.8033146858215332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,8192,0.3250604470570882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,10240,1.1545502344767253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,7168,0.6621804237365723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,8192,0.9449031617906359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,7168,0.27191821734110516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,7168,0.8645093176099988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,6144,0.6411146587795681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,7168,0.8161226908365885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,6144,0.2412186728583442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,6144,0.6637128723992242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,8192,0.88882933722602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,5120,0.44608354568481445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,6144,0.7133830918206109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,5120,0.20739732848273384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,4096,0.3238826592763265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,4096,0.1726764440536499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,3584,0.31844621234469944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,4096,0.4328924549950494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,5120,0.7973724471198188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,5120,0.5985057618882921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,3072,0.2606231106652154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,4096,0.5001715554131402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,3584,0.4075475533803304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,2560,0.21866933504740396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,3072,0.35090043809678817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,3072,0.15103911028967962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,3584,0.4175039927164714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,2048,0.1829679939481947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,2560,0.285791105694241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,3072,0.3709733221266005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,2048,0.11336444483862983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,1536,0.13392178217569986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,2048,0.22768266995747885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,2560,0.3172239992353651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,1024,0.08954311079449123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,2048,0.27023111449347603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,3584,0.19362400637732613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,1536,0.17918488714430067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,768,0.07645510964923434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,1536,0.2012631098429362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,1024,0.14835822582244873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,2560,0.13007467322879368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,512,0.05246577660242716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,768,0.081313775645362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,768,0.12149243884616429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,512,0.08246311214235094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,512,0.07252977954016791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,256,0.03233955634964837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,1536,0.0975724458694458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,512,0.09240177604887222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,128,0.02949777907795376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,256,0.0679324467976888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,256,0.06615110900666979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,1024,0.08384711212582058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,64,0.02470933397610982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,256,0.0654346677992079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,128,0.06514310836791992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,32,0.025387555360794067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,128,0.0615253316031562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,768,0.1036542190445794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,128,0.050085332658555776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,64,0.05950578053792318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,32,0.05860977702670627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,1024,0.13089599874284533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,65536,2.181456036037869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,51200,3.5477831098768444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,65536,4.696546766493055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,51200,1.7205324172973633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,16384,1.0907217661539714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,51200,5.303302341037326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,16384,0.5454008844163682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,16384,1.477420488993327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,51200,5.287756601969401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,12288,0.8427946832444934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,65536,6.877159118652344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,16384,1.648447142706977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,12288,0.39530134201049805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,12288,1.1183448367648654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,65536,6.891957177056207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,10240,0.6529226832919651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,10240,0.3401493231455485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,8192,0.5624426735772027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,10240,0.8993404176500109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,8192,0.2644515567355686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,12288,1.2051608827379015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,8192,0.7336879836188422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,7168,0.5364782015482584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,10240,0.9958595699734158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,7168,0.22850312127007377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,8192,0.7970950868394641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,6144,0.4007262123955621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,6144,0.20018133852216932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,7168,0.6330169041951498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,5120,0.3233395682440864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,7168,0.7005208863152398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,6144,0.7929226557413737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,5120,0.20493155055575898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,128,0.06721244255701701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,6144,0.606279108259413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,4096,0.2577013439602322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,4096,0.14768266677856445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,3584,0.21886755360497367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,4096,0.37743377685546875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,3584,0.15025777286953396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,3584,0.358209768931071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,3072,0.19074843989478218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,4096,0.4073644479115804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,2560,0.16114044189453125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,3584,0.36040888892279727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,3072,0.28322309917873806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,5120,0.572223981221517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,3072,0.32786933581034344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,2560,0.1120871173010932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,2560,0.23830845620897081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,2048,0.14492800500657824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,5120,0.4962933328416612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,1536,0.10964000225067139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,2048,0.10018666585286458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,2560,0.2661271095275879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,2048,0.18842755423651802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,1536,0.0867004460758633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,1536,0.1469386683570014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,2048,0.22048355473412407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,768,0.057370669311947294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,1024,0.07522666454315186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,1536,0.1772773398293389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,3072,0.12933599948883057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,512,0.04282311267322964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,1024,0.126510222752889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,768,0.08483733071221246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,768,0.07366044653786553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,256,0.026517333255873785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,768,0.10008266899320815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,512,0.06531555785073175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,512,0.06663555569118924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,128,0.020830222302012973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,256,0.054639998409483165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,256,0.05973866913053724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,64,0.01853155593077342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,512,0.07895200120078193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,256,0.05468622181150648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,32,0.019124444988038804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,128,0.04826844400829739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,1024,0.07466044690873888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,128,0.041837334632873535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,64,0.04618666569391886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,32,0.045180443260404796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,1024,0.10706222057342529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,128,0.05865955352783203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,65536,1.7960995568169489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,51200,2.420697742038303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,51200,1.3649689356486003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,65536,3.2021270328097873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,16384,0.7814853456285266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,12288,0.5620551109313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,16384,0.4444550938076443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,16384,1.1353848775227864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,51200,3.706879085964627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,12288,0.31316177050272626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,10240,0.4562800195482042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,16384,1.3337084452311199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,51200,4.3026479085286455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,12288,0.8403146531846789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,65536,4.881625705295138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,10240,0.2712079948849148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,12288,0.9965217378404406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,8192,0.36536444558037656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,10240,0.6979893578423394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,8192,0.21223910649617514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,7168,0.31332267655266655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,65536,5.5542479621039496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,7168,0.18516622649298775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,10240,0.8432515462239584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,6144,0.2631288899315728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,8192,0.6622453265719944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,7168,0.6053839789496528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,5120,0.22707022560967338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,6144,0.16266577773623997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,7168,0.583539538913303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,6144,0.42934756808810765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,5120,0.17466844452752006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,5120,0.35604267650180393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,6144,0.49202754762437606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,3584,0.1569582223892212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,4096,0.281896882587009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,4096,0.15150311258104113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,8192,0.5502470864189996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,5120,0.415283547507392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,3072,0.13769066333770752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,3584,0.10874488618638779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,2560,0.1126684480243259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,3584,0.25320090187920463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,3072,0.11598755253685845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,3072,0.2133573293685913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,3584,0.29939021004570854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,2048,0.09516711367501153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,2560,0.09346844752629598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,2560,0.18145155906677246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,3072,0.2611191007826063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,4096,0.18629866176181367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,1536,0.06823911269505818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,2560,0.2196035517586602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,2048,0.08488800128300984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,1024,0.04905422197447883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,2048,0.1509431070751614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,1536,0.07494222455554538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,2048,0.18250844213697645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,1536,0.14228888352711996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,1024,0.08350133233600193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,768,0.03842755489879184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,4096,0.3412346574995253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,512,0.024891555309295654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,1024,0.10397866699430679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,768,0.06536177794138591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,768,0.06572977701822917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,256,0.0174319992462794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,512,0.0576488905482822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,512,0.05062399970160591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,512,0.06417066521114774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,256,0.03995111253526475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,128,0.014514666464593677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,256,0.04355377621120877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,1536,0.11522755357954238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,64,0.012989333106411828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,128,0.03496799866358439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,32,0.013430222041077085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,128,0.041505777173572116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,1024,0.0671280026435852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,128,0.032285332679748535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,64,0.03256355391608344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,32,0.03135555651452806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,768,0.08208177487055461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,256,0.04555111130078634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,51200,2.0455137888590493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,65536,1.7150932947794597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,65536,2.374521679348416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,51200,1.287103123135037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,16384,0.6492488649156358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,16384,0.4440488815307617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,12288,0.42456356684366864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,51200,3.088488048977322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,16384,1.0021742714775932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,10240,0.33266133732265896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,16384,1.1996160083346896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,12288,0.3069760004679362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,12288,0.716020425160726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,65536,4.022247314453125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,8192,0.3041955629984538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,10240,0.26481421788533527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,51200,3.7655618455674915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,12288,0.9001483917236328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,8192,0.20271289348602295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,10240,0.6251813570658366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,7168,0.25380534595913357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,8192,0.4806106885274251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,6144,0.21452000406053331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,65536,4.917466481526692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,7168,0.17951999770270455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,7168,0.4193120002746582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,5120,0.1765448914633857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,8192,0.6120008892483181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,6144,0.1616097821129693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,4096,0.14929778046078152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,5120,0.1364533371395535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,7168,0.5218213399251302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,5120,0.3042159875233968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,6144,0.46567201614379883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,4096,0.11672177579667832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,3584,0.1292568842569987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,4096,0.24942133161756727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,3072,0.11335199409061009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,3584,0.11087377866109212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,10240,0.7416444354587131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,4096,0.30784445338779026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,3584,0.21523733933766684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,2560,0.09395911296208699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,3072,0.09822755389743382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,3584,0.26881334516737193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,6144,0.3607929017808702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,2048,0.0707119968202379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,2560,0.15780799918704563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,2560,0.09097599983215332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,3072,0.23809422387017143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,1536,0.060544888178507485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,2048,0.08306666877534655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,2048,0.12837066915300158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,2560,0.2006631162431505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,1024,0.04053333401679993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,1536,0.07364444600211249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,1536,0.1001617776023017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,2048,0.16371377309163412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,5120,0.3863946596781413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,1024,0.07163021961847942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,1024,0.0657164454460144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,1536,0.13012622462378606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,512,0.019651555352740817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,768,0.056353780958387584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,768,0.0644177794456482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,1024,0.0935751133494907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,256,0.014501333236694336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,768,0.0742808911535475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,3072,0.18598311477237275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,512,0.05079999897215101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,512,0.04281155599488152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,128,0.01197866681549284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,256,0.03746666510899862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,512,0.055883553293016225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,64,0.010671999719407825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,128,0.028296000427669946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,128,0.03595733311441209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,32,0.01108977778090371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,256,0.03785955574777391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,128,0.027801778581407335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,64,0.02586577832698822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,768,0.03045066528850132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,32,0.024359110328886244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,256,0.03257955445183648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,51200,1.3307502534654405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,65536,1.6021173265245225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,65536,1.346226692199707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,16384,0.41272889243231875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,16384,0.33120089107089573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,16384,0.786493354373508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,12288,0.2683493296305339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,51200,2.533209694756402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,51200,1.0522729025946724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,12288,0.24608887566460502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,16384,1.0460533565945096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,12288,0.591220431857639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,10240,0.21393777264489067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,10240,0.5088257789611816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,8192,0.17568800184461805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,10240,0.6436986923217773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,51200,3.255465825398763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,10240,0.24749777052137586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,8192,0.4016995429992676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,7168,0.177419556511773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,8192,0.16200444433424208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,6144,0.1479840013715956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,8192,0.5281093385484483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,7168,0.1437848938835992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,7168,0.34969154993693036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,12288,0.7831875483194987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,5120,0.11615999539693196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,6144,0.1253688865237766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,7168,0.4584355354309082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,6144,0.32526132795545787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,5120,0.10851733552085029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,4096,0.10608177714877659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,5120,0.25295289357503253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,6144,0.39548089769151473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,4096,0.09474133120642768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,3584,0.09210666682985093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,4096,0.20659022861056855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,3072,0.06849510802163018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,3584,0.0878515574667189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,3584,0.17966755231221518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,4096,0.2660577827029758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,2560,0.06637244754367404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,3072,0.08064177963468763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,3584,0.23727732234530977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,3072,0.20737600326538086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,2560,0.1325368881225586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,2048,0.053915556934144765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,2560,0.17580978075663248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,2048,0.10850133498509724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,5120,0.33103378613789874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,1536,0.04179111123085022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,2048,0.06930133369233873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,2048,0.14469156000349256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,1536,0.06328355603747897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,1024,0.02641244398223029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,1536,0.0836168858740065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,3072,0.15604266855451795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,768,0.019834667444229126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,1536,0.11245866616566975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,65536,3.268645392523872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,1024,0.05727822250790066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,1024,0.06005866660012139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,512,0.014736889137162102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,2560,0.07593866851594713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,768,0.046875556310017906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,768,0.05057333244217766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,512,0.03266400098800659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,256,0.010334221853150262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,512,0.03530577818552653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,128,0.008637333081828224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,768,0.06391200092103747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,256,0.03013422091801961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,512,0.0483528905444675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,256,0.025927111506462097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,64,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,32,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,128,0.02160888910293579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,128,0.029841777351167467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,256,0.03198400139808655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,64,0.019631110959582858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,32,0.018194667167133756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,1024,0.08140444755554199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,65536,0.9154515796237521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,128,0.024361777636739943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,51200,0.6784239874945747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,65536,1.3061111238267686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,16384,0.20526311132642958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,65536,2.601928922865126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,51200,1.0110267003377278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,16384,0.6688790851169162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,16384,0.3028586705525716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,12288,0.15798400508032903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,51200,2.0355146196153426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,10240,0.13986755741967097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,12288,0.24861600663926867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,12288,0.4844542079501682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,65536,3.6656888326009116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,16384,0.9142061869303385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,10240,0.406988435321384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,8192,0.10626577668719822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,10240,0.20771200127071807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,12288,0.6894222365485297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,51200,2.8489422268337674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,8192,0.15472888946533203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,8192,0.3273048930697971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,10240,0.5794293085734049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,6144,0.0834346678521898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,7168,0.13682133621639675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,7168,0.28561422559950084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,5120,0.07467644744449191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,8192,0.4573369026184082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,6144,0.12064177460140652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,6144,0.24723643726772734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,4096,0.05379555622736613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,7168,0.4006391101413303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,5120,0.10505421956380208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,5120,0.2066826687918769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,6144,0.3486817677815755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,3584,0.053620444403754346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,7168,0.09655733240975274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,4096,0.0918240017361111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,4096,0.16748089260525176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,3072,0.04522488845719231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,5120,0.28922756512959796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,3584,0.08426133129331802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,3584,0.1475084490246243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,2560,0.03532000051604377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,3072,0.07806311051050822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,4096,0.2345813380347358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,3072,0.12751466698116726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,2048,0.02736888825893402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,3584,0.20946310626135933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,2560,0.07301421960194905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,2560,0.10843822028901841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,1536,0.021478222476111516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,3072,0.18278400103251138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,2048,0.06684711244371203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,2048,0.08910489082336426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,2560,0.15497155984242758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,1536,0.06138044595718384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,2048,0.12542222605811224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,1536,0.06769688924153645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,768,0.011295110815101199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,1024,0.04996088809437222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,1024,0.047931555244657725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,1536,0.09895022047890557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,512,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,768,0.03565422362751431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,768,0.037441776858435735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,1024,0.07096711132261488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,256,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,512,0.031207998593648274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,512,0.02751733362674713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,768,0.05408089028464424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,128,0.006235555642180973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,256,0.029703110456466675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,256,0.01923199991385142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,512,0.040082666609022356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,64,0.005670222143332164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,32,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,256,0.02572800053490533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,128,0.029495110114415485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,128,0.015358222855461968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,64,0.01313511116637124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,128,0.01995644470055898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,1024,0.014324444863531323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,32,0.01203555530971951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,65536,0.6857048670450846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,51200,0.5526409149169922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,16384,0.1847155491511027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,65536,1.3162737952338326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,16384,0.6117235819498698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,51200,1.7841057247585719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,51200,1.0195919672648113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,12288,0.13226932949490017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,16384,0.3120346599155002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,65536,2.2818692525227866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,12288,0.2401199870639377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,12288,0.4460942480299208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,10240,0.11561332808600532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,10240,0.3664284547170003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,10240,0.21369866530100504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,8192,0.090137779712677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,8192,0.2957715458340115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,8192,0.15658756097157797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,7168,0.08175022072262235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,7168,0.2577022181616889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,65536,4.230377621120877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,7168,0.14260533120897081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,6144,0.07103821966383192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,6144,0.22425511148240831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,10240,0.5388408766852485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,6144,0.1214853392706977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,5120,0.06047999858856201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,16384,0.8520248730977377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,12288,0.6423350969950358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,5120,0.18654044469197592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,4096,0.049301332897610135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,5120,0.10559999942779541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,5120,0.27599112192789715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,4096,0.1511039998796251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,3584,0.0436062216758728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,4096,0.0922737783855862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,51200,2.637173334757487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,4096,0.22416533364189994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,3584,0.08452444606357151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,3072,0.037596444288889565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,3584,0.13362044758266872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,3584,0.19803555806477866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,3072,0.07816444502936469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,3072,0.11551288763682048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,2560,0.032792889409595065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,3072,0.17187110582987467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,2560,0.07243111398484972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,8192,0.43503554662068683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,2048,0.026015111141734656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,2560,0.09755911429723103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,1536,0.01902222302224901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,2048,0.06598844130833943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,2560,0.14558488792843288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,2048,0.07969066831800672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,1024,0.00903555585278405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,1536,0.06020622120963203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,2048,0.11971467071109348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,1024,0.04456444581349691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,768,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,1024,0.06383022334840563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,768,0.03237688872549269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,768,0.03386755453215705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,512,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,768,0.04923466510242886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,7168,0.3763119909498427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,512,0.02364088926050398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,256,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,512,0.031199110878838435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,1024,0.0418906675444709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,512,0.036107556687461004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,256,0.015832000308566622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,128,0.00423911131090588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,256,0.029878222280078467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,256,0.023009777069091797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,128,0.012081777883900536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,128,0.029501332177056208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,32,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,64,0.010044444766309526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,32,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,1536,0.05985333522160848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,1536,0.09187022182676528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,65536,0.6386657820807563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,51200,0.49666934543185765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,64,0.004545777622196409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,16384,0.16380000114440918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,12288,0.12456622388627793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,10240,0.10241955518722534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,12288,0.454819573296441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,16384,0.5847617785135905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,8192,0.08671644661161636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,10240,0.3566177686055501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,7168,0.07330755392710368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,6144,0.06300889121161567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,8192,0.2901279926300049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,7168,0.2535591125488281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,5120,0.0540648897488912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,65536,2.2185351053873696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,51200,1.7234978146023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,4096,0.04531822270817227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,6144,0.21932710541619194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,3584,0.03841244512134128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,3072,0.033187554942237005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,2560,0.02661955522166358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,5120,0.18253600597381592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,4096,0.14835555023617214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,3584,0.13061511516571045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,2048,0.02091466718249851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,1536,0.013557333085272046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,1024,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,2560,0.09438844521840413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,1536,0.0573271115620931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,2048,0.07620977693133883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,768,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,512,0.005528000079923206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,768,0.03086844417783949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,256,0.004243555582231945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,256,0.014463111758232117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,6144,0.3282257715861003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,128,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,64,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,3072,0.11286755402882893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,32,0.004854222138722737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,128,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,64,0.008964444200197855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,32,0.009878222313192155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,1024,0.039453334278530545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,512,0.022234666678640578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,16384,0.170232892036438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,65536,0.7225271330939399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,51200,0.572090678744846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,12288,0.1256355577045017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,10240,0.10536799828211467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,8192,0.08560444249047174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,16384,0.56222046746148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,12288,0.429307566748725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,10240,0.3560097747378879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,7168,0.07638933261235555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,6144,0.06542133622699313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,8192,0.2887795501285129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,5120,0.056052446365356445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,4096,0.04549244377348158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,7168,0.253311104244656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,6144,0.2175920009613037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,3584,0.04057333204481337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,5120,0.18203377723693848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,3072,0.03559822175237868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,4096,0.14803022808498806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,2560,0.03012088934580485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,65536,2.217588424682617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,2048,0.024783111280865137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,1536,0.014681778020328946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,3584,0.13305511739518908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,1024,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,3072,0.11444977919260661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,768,0.005863111052248213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,2560,0.09615377585093181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,512,0.00453333349691497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,2048,0.07857244544559054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,1536,0.06003822220696343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,256,0.003552888830502828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,128,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,1024,0.04185511006249321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,768,0.032588445478015475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,512,0.02323644525474972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,32,0.004192000048028098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,256,0.014452444182501899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,128,0.010407110883129967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,64,0.009665778113736046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,51200,1.7235707177056208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,32,0.010006222460005019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,64,0.00254488881263468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,65536,3.3712747361924915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,16384,20.670817057291668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,51200,27.010932074652775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,16384,8.349325391981337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,128,0.018918222851223417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,16384,29.42264472113715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,16384,42.64460923936632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,12288,6.593504163953994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,12288,18.64190673828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,12288,21.984198676215275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,10240,13.656807793511284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,12288,30.939497205946182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,10240,5.355158911810981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,10240,18.75591532389323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,51200,66.80838351779514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,10240,25.640672471788193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,8192,13.117453681098091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,8192,4.2706116570366754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,7168,3.581256866455078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,8192,14.551389906141493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,7168,11.019840664333769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,8192,22.001458062065975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,7168,12.99029541015625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,6144,9.791414048936632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,6144,3.1393758985731335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,7168,17.566870795355904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,6144,10.609001159667969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,6144,12.509596082899305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,5120,2.6138223012288413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,5120,6.914014180501302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,5120,8.43822987874349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,4096,2.086035622490777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,5120,7.6655061509874125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,4096,5.637082841661241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,4096,5.876681857638889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,3584,1.846385743882921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,4096,5.7186330159505205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,3584,5.806885189480251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,3584,5.220255109998915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,3072,3.821862962510851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,3584,4.859568701850043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,51200,93.4251708984375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,3072,4.30955547756619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,3072,1.6950782140096028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,3072,4.132748497856988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,2560,1.3965280320909288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,2560,3.1823387145996094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,2560,3.4388703240288625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,2048,1.134319093492296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,2560,3.4436009724934897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,2048,2.9659991794162326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,2048,2.821279101901584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,1536,1.9315910339355469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,2048,2.795155631171333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,1536,0.8713359832763672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,1024,1.4828755060831706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,1024,0.6284106572469076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,1536,2.226994620429145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,768,1.1743591096666124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,1536,2.221143934461806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,1024,1.6189599566989477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,768,0.5713191032409668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,1024,1.6336008707682292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,512,0.5343822373284234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,768,1.3208391401502821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,512,0.8987911012437609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,256,0.5340133243136936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,768,1.290228419833713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,512,1.0092248916625977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,512,0.962858623928494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,128,0.5483182271321615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,256,0.6341350873311361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,128,0.42923733923170304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,256,0.8999822404649523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,256,0.7141831186082629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,64,0.43129600418938535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,128,0.8625920083787707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,128,0.5425093438890245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,32,0.37021599875556094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,64,0.8803955713907877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,32,0.9869769414265951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,51200,127.36234537760417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,65536,26.308490329318577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,51200,21.06408182779948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,16384,17.595833672417534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,51200,53.735799153645836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,16384,6.8744905259874125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,16384,22.947650485568577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,65536,70.06402587890625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,16384,33.590494791666664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,12288,13.48792945014106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,12288,5.184761895073785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,51200,73.87545437282986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,10240,12.05946011013455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,12288,17.781053331163196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,10240,4.4111713833279085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,12288,24.170186360677082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,10240,14.529686821831596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,8192,9.833752102322048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,10240,20.05565897623698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,8192,3.2762097252739797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,51200,99.9231906467014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,7168,2.884951909383138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,8192,11.53345913357205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,7168,8.10229237874349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,8192,17.265767415364582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,7168,10.069208780924479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,6144,6.803154839409722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,6144,2.464697731865777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,7168,13.705262925889757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,6144,8.346684773763021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,6144,9.87003411187066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,5120,2.066991170247396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,5120,5.603241814507379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,5120,6.432153065999349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,5120,5.997925228542751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,4096,1.626052432590061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,4096,4.856064690483941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,4096,4.7152993943956165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,4096,4.4399189419216585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,3584,3.9418784247504344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,3584,1.4314257303873699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,3584,3.979864756266276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,3584,3.8366160922580295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,3072,1.2408249113294814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,3072,3.421653323703342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,3072,3.250991185506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,2560,1.0336062113444011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,2560,2.8778711954752603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,3072,3.2941218482123484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,2560,2.709048800998264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,2048,0.8356569078233508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,2048,1.938112046983507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,2560,2.7625111473931203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,1536,1.5523004531860352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,1536,0.6422862476772732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,2048,2.2060835096571183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,2048,2.2195538414849176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,1024,1.0407564375135634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,1536,1.6819137997097438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,1024,0.554857783847385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,1024,1.2362826665242512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,1024,1.2461279763115776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,768,0.8301359812418619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,1536,1.7288097805447047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,768,0.43299823337131077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,512,0.8294142087300619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,768,1.0027875900268555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,512,0.7713502248128256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,512,0.7250933117336698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,512,0.5849617852105035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,256,0.32195109791225857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,256,0.6951271163092719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,128,0.44652001063028973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,256,0.56186490588718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,128,0.33322400516933864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,64,0.2729102240668403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,128,0.6631280051337348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,128,0.42581955591837567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,32,0.2862871223025852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,768,1.0178213119506836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,64,0.6823333104451498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,32,0.7718888918558756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,256,0.4453893237643772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,65536,133.76216634114584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,65536,10.587254842122396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,65536,96.56414794921875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,51200,16.217880249023438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,51200,8.244288974338108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,16384,6.103808932834202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,65536,21.973620944552952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,16384,2.6140374077690973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,12288,4.343042585584852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,16384,10.84510972764757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,51200,24.370858086480037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,16384,7.811063130696614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,12288,1.9033084445529516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,12288,7.851396348741319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,10240,3.73358154296875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,12288,5.614073859320746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,10240,1.5704640282524958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,65536,31.211496988932293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,51200,32.66004774305556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,8192,1.0092106925116646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,10240,4.577320098876953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,10240,6.484946780734592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,8192,3.135413275824653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,7168,2.5217723846435547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,7168,0.8612817658318414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,8192,3.717375013563368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,6144,2.1299911075168185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,6144,0.8163821962144641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,8192,5.531182183159722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,7168,3.3467831081814237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,6144,2.754667494032118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,7168,4.429524315728082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,6144,3.0555778079562717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,5120,0.731204456753201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,5120,1.8693404727511937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,5120,1.9779190487331813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,4096,0.48934576246473527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,5120,2.08823733859592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,4096,1.3855777316623266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,4096,1.441627608405219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,3584,1.2490293714735243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,3584,0.5081964598761665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,4096,1.4792924457126195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,3072,0.4294488959842258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,3584,1.1832328372531467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,3584,1.2238978279961479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,3072,1.06696351369222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,3072,0.9909511142306857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,2560,0.3233155674404568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,3072,1.0398231082492404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,2560,0.8692284690009223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,2560,0.8475262324015299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,2048,0.6768942409091525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,2048,0.2682524522145589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,2048,0.6541848712497288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,2560,0.8656195534600152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,1536,0.215155561765035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,1536,0.5616870986090766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,1024,0.3519582218594021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,1536,0.5221279991997613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,1024,0.39728887875874835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,1024,0.17149155669742158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,768,0.2957368956671821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,1024,0.3759644561343723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,512,0.2233511077033149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,768,0.1504533290863037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,768,0.33077777756585014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,512,0.1300168832143148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,256,0.14470844798617893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,512,0.25874310069613987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,768,0.30774132410685223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,128,0.11268177297380234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,256,0.1190604501300388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,512,0.3591582245296902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,65536,45.29454549153646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,256,0.1939457787407769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,2048,0.8237537807888455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,128,0.22063377168443468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,64,0.08994755480024551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,32,0.093304885758294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,128,0.12140622403886582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,64,0.2230595482720269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,1536,0.5162089135911729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,32,0.2520071135626899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,256,0.22903998692830405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,65536,8.056890699598524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,51200,12.029847039116753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,51200,6.081999037000869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,16384,3.8402197096082897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,65536,16.686250474717884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,16384,1.836666742960612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,51200,18.208082411024304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,12288,3.326463063557943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,16384,5.7205005221896705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,16384,8.150904337565104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,12288,1.1103244357638888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,65536,23.87423027886285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,12288,5.8757811652289496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,10240,3.153437296549479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,51200,24.453965928819443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,12288,4.438064999050565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,10240,0.9523457421196831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,8192,0.7645244598388672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,8192,2.335648854573568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,10240,4.848384009467231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,10240,3.6227874755859375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,8192,3.0223511589898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,8192,2.9219529893663196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,7168,1.9094195895724828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,7168,0.8234320216708713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,6144,0.5043493376837836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,7168,2.1660452948676214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,6144,1.7505938212076824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,7168,2.805615954928928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,128,0.14440800084008112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,6144,1.6712160110473633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,5120,1.3953172895643446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,6144,1.883252461751302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,5120,0.5238489045037164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,4096,1.0021564695570204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,5120,1.2816230985853407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,4096,0.5273608631557888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,5120,1.4224782519870336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,4096,0.9983563952975802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,3584,0.8939599990844727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,65536,33.43328857421875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,3584,0.40913955370585126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,4096,1.0315831502278645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,3072,0.28112443288167316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,3584,0.918576028611925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,3072,0.8369573487175835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,3072,0.7467679977416992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,2560,0.6483608881632487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,2560,0.2482008934020996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,2048,0.5510497623019749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,2560,0.6950897640652127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,3072,0.9504275851779513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,2048,0.20991022057003447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,2048,0.5162524647182888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,2560,0.6274684270222982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,1536,0.17172533935970732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,1536,0.44154932763841415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,3584,0.9388568666246203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,1024,0.27043554517957896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,1536,0.4139333301120334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,2048,0.5927466816372341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,768,0.21105066935221353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,1024,0.19086666901906332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,1024,0.3263297875722249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,1536,0.41564000977410215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,512,0.14472444852193198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,768,0.11958400408426921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,768,0.3117662270863851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,1024,0.29880889256795246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,512,0.10616978009541829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,128,0.083405335744222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,512,0.300817781024509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,256,0.09555288818147446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,256,0.17848800288306343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,512,0.1942382256189982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,64,0.06938578022850884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,128,0.09549688630633885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,32,0.07111022207472059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,128,0.1621262232462565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,128,0.11151377360026042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,64,0.16304088963402644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,32,0.1785608927408854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,256,0.1054711143175761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,768,0.23796889517042372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,256,0.15420088503095838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,51200,5.096719106038411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,65536,6.646772596571181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,16384,3.507224824693468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,51200,12.728655497233072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,16384,6.891520182291667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,65536,14.355445014105904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,16384,1.5141706466674805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,16384,4.9160410563151045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,12288,2.484016842312283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,12288,0.9960098266601562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,51200,15.923663669162325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,10240,2.5178915659586587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,12288,3.5710843404134116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,10240,0.8367395401000977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,12288,4.875602722167969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,10240,3.421899583604601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,10240,2.8749910990397134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,65536,19.948326958550346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,51200,20.621741400824654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,8192,0.5807315508524576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,8192,1.7480204900105794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,7168,0.5832711325751411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,8192,2.021086162990994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,7168,1.6291804843478734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,6144,1.4187795850965712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,8192,2.5496514638264975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,6144,0.5178808636135525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,7168,1.6628008948432074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,7168,1.9927484724256728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,6144,1.3515324062771266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,5120,0.3654915491739909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,5120,1.045630243089464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,6144,1.7181982464260523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,5120,1.0996195475260417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,4096,0.9472133848402234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,4096,0.34190932909647626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,4096,0.8469573126898872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,5120,1.2058853573269315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,3584,0.2959368758731418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,4096,0.8695128758748373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,3584,0.865598201751709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,3584,0.7478186819288465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,3072,0.2584497663709852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,3072,0.7036648856268989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,3584,0.7202897601657443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,3072,0.7299662166171603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,2560,0.28680088784959584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,2560,0.6295813454522027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,3072,0.6217981974283854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,2560,0.5344302389356825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,2048,0.39354756143358016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,2048,0.2395697699652778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,2560,0.5315324465433756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,1536,0.3214142322540283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,2048,0.4413111209869385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,1024,0.22243555386861166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,1536,0.14446400271521673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,1536,0.3425191243489583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,2048,0.4439520041147868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,768,0.17330488893720838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,1024,0.13850843906402588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,1024,0.37749067942301434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,1536,0.3650800122155084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,512,0.12921422057681614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,768,0.10611110925674438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,1024,0.24812711609734428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,768,0.2013875510957506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,256,0.08988710906770493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,512,0.09888088703155518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,768,0.2027742200427585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,512,0.1920631064309014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,256,0.08568088875876533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,128,0.07611378033955891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,256,0.14484266440073648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,512,0.16236444314320883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,64,0.06192355685763889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,128,0.08562044302622478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,32,0.06041510899861654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,128,0.13692088921864828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,128,0.09578577677408855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,64,0.1373466650644938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,32,0.147507561577691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,65536,28.352618747287327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,256,0.12473244137234157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,51200,4.045515696207683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,65536,5.2082540724012585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,51200,8.147218492296007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,16384,2.815980487399631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,16384,1.0169964896308052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,16384,4.722724490695529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,65536,11.256096733940971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,51200,10.284745110405817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,16384,3.252052519056532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,12288,2.215728971693251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,12288,0.7303928799099393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,12288,2.4783590104844833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,10240,2.0803760952419705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,12288,3.100735134548611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,10240,2.1132142808702254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,10240,0.6427706612481011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,10240,1.8826062944200304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,65536,13.250370449490019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,8192,0.46862930721706814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,8192,1.5042070812649195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,8192,1.4443608389960394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,8192,1.4573013517591689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,7168,0.3975200123257107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,7168,1.3289413452148438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,7168,1.1928400463528104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,51200,16.51187981499566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,7168,1.2486480077107747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,6144,0.3552533255683051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,6144,1.065950181749132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,6144,1.0827129152086046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,5120,0.8706746631198459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,5120,0.35825422075059676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,6144,1.0661982430352104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,5120,0.885658688015408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,4096,0.2508702278137207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,4096,0.6868835555182563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,5120,0.8827066421508789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,4096,0.7284711201985677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,3584,0.6316586600409614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,3584,0.2233875592549642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,4096,0.6766728825039334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,3584,0.6157520082261827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,3072,0.20209956169128418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,3072,0.5007413228352865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,3584,0.5871031019422743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,3072,0.5228648715549046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,2560,0.18170489205254448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,2048,0.3356195555792914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,2560,0.4380631181928847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,3072,0.6977359983656141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,2560,0.432845327589247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,2048,0.16377154986063638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,2048,0.35748889711168075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,1536,0.27151645554436576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,1024,0.18918400340610078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,1536,0.13349600632985434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,2048,0.357859558529324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,1536,0.29248889287312824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,1024,0.10142844253116184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,2560,0.4366275469462077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,1536,0.2809608777364095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,768,0.16938755247328016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,768,0.08681777450773452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,1024,0.20959199799431694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,512,0.07939110861884223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,768,0.1796097755432129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,512,0.13407288657294378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,512,0.13780889246198866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,256,0.11803200509813096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,128,0.05605600277582804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,256,0.07168889045715332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,256,0.10416355397966172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,768,0.1440755526224772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,64,0.04654666781425476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,128,0.07165600193871392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,128,0.11055999994277954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,32,0.04793333344989353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,512,0.11209332942962646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,128,0.07995200157165527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,64,0.1114728848139445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,32,0.11441955301496719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,256,0.06930133369233873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,65536,22.425828721788193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,1024,0.22967110739813912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,51200,3.4336213005913625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,65536,4.627911885579427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,16384,2.6480178833007812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,16384,3.812355465359158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,51200,9.287104288736979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,51200,9.041692097981771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,16384,0.880360868242052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,65536,10.800199720594618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,16384,2.8723661634657116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,12288,0.597309324476454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,65536,11.741503397623697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,10240,1.6926790873209636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,12288,2.412503984239366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,12288,1.999452379014757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,12288,2.1063014136420355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,10240,0.6213697857326931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,10240,1.7458889219495985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,8192,0.41541777716742617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,51200,13.747612847222221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,8192,1.433311144510905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,8192,1.2331973181830513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,10240,1.6837839550442164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,8192,1.248670260111491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,7168,1.2732044855753581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,7168,0.4254008928934733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,6144,0.9288382000393338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,6144,0.467299567328559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,7168,1.1132755279541016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,7168,1.086949348449707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,6144,0.9184755749172635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,5120,0.27289067374335396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,6144,0.918267567952474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,4096,0.6354853312174479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,4096,0.2308551205529107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,5120,0.830892456902398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,5120,0.7231804529825846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,4096,0.5931795438130697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,3584,0.572856002383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,5120,0.7548186514112684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,4096,0.5945990880330404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,3584,0.20372622542911104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,3584,0.5529466734992133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,3072,0.18062310748630098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,3584,0.5146489143371582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,3072,0.48969512515597874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,2560,0.3137884404924181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,3072,0.454603566063775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,2560,0.2012711101108127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,3072,0.4634355439080133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,2560,0.3870515558454726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,2048,0.13971288998921713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,2048,0.32899732059902614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,1536,0.23388976520962187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,2560,0.38541600439283585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,1024,0.15354843934377035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,1536,0.10938577519522773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,2048,0.32358045048183864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,1536,0.25746221012539333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,768,0.12791377968258327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,1024,0.08777511119842529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,1024,0.19800089465247261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,768,0.08187822500864665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,2048,0.28248267703586155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,512,0.09519377681944106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,768,0.14798043833838567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,768,0.15324355496300593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,512,0.072488890753852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,256,0.0665466652976142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,512,0.1191520028644138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,128,0.049118220806121826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,256,0.06618399752510919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,512,0.1248755587471856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,65536,19.540079752604168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,256,0.10417511065800984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,64,0.04605866803063286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,128,0.06590488884184095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,32,0.04900888933075798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,1536,0.24799556202358672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,128,0.09762577878104316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,64,0.09860977861616348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,128,0.07143821981218126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,1024,0.18731910652584502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,32,0.10093510813183254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,256,0.09338133202658759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,65536,3.6614409552680125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,51200,2.9233396318223743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,16384,2.2720959981282554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,51200,6.816871219211155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,16384,0.693843576643202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,16384,3.3543999989827475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,51200,7.751005384657119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,65536,8.572660658094618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,16384,2.4271119435628257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,12288,0.5208693080478245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,12288,1.7038764953613281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,12288,1.7631413141886394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,10240,0.4421537717183431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,12288,1.777929729885525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,10240,1.4892862108018663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,65536,10.139827304416233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,10240,1.4696328904893663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,8192,0.40552979045444065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,10240,1.444767951965332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,8192,1.1414737701416016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,8192,1.0747831132676866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,7168,1.0356658299763997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,7168,0.39229599634806317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,8192,1.3014098273383248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,51200,12.536064147949219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,6144,0.869801786210802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,7168,0.9546000162760416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,6144,0.26972712410820854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,5120,0.610605345831977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,5120,0.23306666480170357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,6144,0.941832012600369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,6144,0.8114346928066678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,4096,0.6094417572021484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,5120,0.9529999627007378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,5120,0.6531297895643446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,7168,0.9840933481852213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,4096,0.5613199869791666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,4096,0.5174044503106011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,3584,0.17918044990963408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,3584,0.5392604404025608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,3072,0.39507378472222227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,3584,0.46052000257703996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,3072,0.1940257814195421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,3584,0.4615955352783203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,3072,0.41095556153191465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,2560,0.34195378091600204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,4096,0.20121778382195366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,2560,0.1399644480811225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,2560,0.337299558851454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,2048,0.25399910079108345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,1536,0.19501245021820068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,2048,0.11889155705769856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,2048,0.3935448858473036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,2560,0.43439555168151855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,1536,0.2371137672000461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,1024,0.1347973346710205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,1536,0.10586933294932048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,2048,0.28015732765197754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,768,0.10688800281948513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,1024,0.08518399794896443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,1536,0.22484087944030762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,1024,0.16509777969784206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,512,0.08061333497365315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,768,0.07378488779067993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,1024,0.17372088962131074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,3072,0.3982577853732639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,256,0.051811556021372475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,768,0.13701600498623318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,512,0.0678408874405755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,512,0.1045902238951789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,128,0.039397332403394915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,65536,16.725024753146702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,256,0.059827552901373975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,256,0.09075022406048244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,64,0.03386577632692125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,128,0.05969599882761637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,128,0.08484444353315566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,256,0.08245511187447442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,32,0.035437332259284124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,128,0.06299466556972928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,64,0.08493510882059734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,32,0.08774044116338094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,768,0.12795821825663248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,512,0.10931999815834893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,51200,2.354567209879557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,65536,3.1938141716851125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,16384,1.7523804770575628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,16384,2.2881751590304904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,51200,6.163623809814453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,16384,0.5839689042833116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,65536,7.552641974555121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,16384,2.1148933834499783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,12288,1.3934941821628148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,51200,6.5720409817165795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,12288,0.44834576712714297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,12288,1.4495644039577906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,10240,1.3113173378838432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,65536,8.727648417154947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,12288,1.4557306501600478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,10240,1.1484115388658311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,10240,0.3808017836676703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,8192,0.30743466483222115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,10240,1.1383191214667427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,8192,0.9614426294962565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,7168,0.8195119963751899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,8192,0.915687984890408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,8192,0.8960071139865451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,7168,0.32689865430196124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,7168,0.7951439751519097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,6144,0.24306845664978027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,6144,0.6937075720893012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,7168,0.7674604521857368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,5120,0.541993776957194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,6144,0.6672568851047092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,5120,0.21052889029184976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,6144,0.6657475365532769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,4096,0.4481164614359538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,5120,0.5977199872334799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,51200,10.736244201660156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,4096,0.17270755767822266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,4096,0.4471217791239421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,5120,0.6073750919765896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,3584,0.38606754938761395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,3584,0.1536346673965454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,4096,0.4485413233439128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,3072,0.3416435453626845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,3584,0.41398223241170246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,3072,0.1369075510236952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,2560,0.26525510682000053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,65536,13.100800408257378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,3072,0.34598400857713485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,3584,0.39780534638298887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,2560,0.1210506624645657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,2048,0.22398045327928331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,2560,0.32163911395602757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,2048,0.10463644398583306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,1536,0.16786489221784803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,1024,0.11533955732981364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,2560,0.292614221572876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,2048,0.2946675618489583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,1536,0.08548000123765732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,1536,0.2143075598610772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,768,0.09143022033903335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,2048,0.2421626779768202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,1024,0.07211199733946058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,1536,0.1915768914752536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,512,0.06496177779303657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,1024,0.14933510621388754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,768,0.06565866867701213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,768,0.11317333910200332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,256,0.04253066579500834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,1024,0.1441004408730401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,512,0.060989333523644336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,512,0.08914044168260361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,768,0.1179271141688029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,128,0.03676533367898729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,256,0.05494222376081678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,256,0.0795253316561381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,512,0.09690399964650472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,64,0.032857778999540545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,256,0.07106844584147136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,128,0.05492266681459215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,3072,0.41518934567769367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,64,0.07181511322657268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,128,0.05474844574928284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,32,0.07281511359744601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,32,0.03395377927356296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,128,0.07177422444025676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,65536,2.3773714701334634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,16384,1.3410950766669378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,51200,4.108265770806207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,51200,1.969100528293186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,16384,0.6145013173421224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,65536,5.931574927435981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,16384,1.693282657199436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,12288,0.9812675052218967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,16384,1.71998045179579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,12288,0.4048524432712131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,51200,5.42357677883572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,10240,0.8193662431504992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,12288,1.1993012958102756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,65536,7.050209045410156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,12288,1.189368036058214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,10240,0.37879467010498047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,8192,0.595594671037462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,10240,0.9562533696492513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,8192,0.2633662223815918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,10240,0.9655137591891818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,8192,0.7719262440999349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,7168,0.5937600135803223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,7168,0.22934932178921172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,8192,0.7286097738477918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,7168,0.648413340250651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,6144,0.5528844197591146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,6144,0.19945066505008271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,7168,0.6485590934753418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,5120,0.4266346560584174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,6144,0.6416568756103516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,51200,8.74587927924262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,5120,0.17190844482845732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,6144,0.5629430876837836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,4096,0.36576178338792587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,4096,0.14410400390625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,3584,0.2760008970896403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,5120,0.7662026617262098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,5120,0.46834664874606663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,4096,0.3731555673811171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,3584,0.13448533746931288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,3072,0.260247098075019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,3584,0.44310845269097227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,4096,0.39867644839816624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,2560,0.2165964444478353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,3072,0.11731999450259739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,3584,0.33633867899576825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,2048,0.16730133692423502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,3072,0.2928568787044949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,2560,0.12682667043473986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,2560,0.2573537826538086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,1536,0.13618932829962835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,2048,0.0887297789255778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,2560,0.2577013439602322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,2048,0.19785777727762857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,1024,0.09246933460235596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,1536,0.07384444607628717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,1536,0.15672088993920222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,2048,0.208278218905131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,768,0.06934755378299289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,1024,0.06195111407174004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,1024,0.11531644397311741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,1536,0.16374133692847356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,512,0.055148445897632174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,65536,11.1991458468967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,768,0.05828266673617893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,768,0.09177244371838039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,256,0.03330755564901564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,1024,0.1248755587471856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,3072,0.2850426567925347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,128,0.02924622098604838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,512,0.052595555782318115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,768,0.10043733649783665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,512,0.07524889045291476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,256,0.0479964448346032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,64,0.02548888822396596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,32,0.025891555680169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,256,0.06355733341640897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,128,0.047516445318857826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,128,0.05888711081610786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,256,0.060306668281555176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,64,0.058838221761915416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,128,0.0461662213007609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,32,0.05915111303329468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,512,0.08043822315004137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,65536,2.1983040703667536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,51200,4.242927127414279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,65536,4.738106621636285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,16384,1.3503342734442816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,51200,4.764806959364149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,16384,0.4983226458231608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,51200,1.5804595947265625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,16384,1.404527982076009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,12288,1.0233760409884982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,16384,1.5310017267862956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,12288,0.34066134028964573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,65536,6.281125386555989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,51200,6.721591949462891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,12288,1.14501953125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,12288,1.0313724941677518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,10240,0.32772088050842285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,10240,0.908938619825575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,8192,0.23107732666863334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,10240,0.8603813383314344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,10240,0.8609297540452746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,7168,0.6074390941196018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,7168,0.20417955186631945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,65536,8.787229749891493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,7168,0.587407112121582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,6144,0.5087599754333496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,7168,0.6042666435241699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,8192,0.6824124654134115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,6144,0.5288809140523275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,5120,0.4256426758236355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,8192,0.6544533305697972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,6144,0.5111404524909126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,5120,0.15459733539157444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,5120,0.4194168779585097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,4096,0.3341831101311578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,8192,0.6833484437730578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,5120,0.4264693260192871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,4096,0.1296373340818617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,3584,0.2886417706807454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,4096,0.3289555443657769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,3584,0.11911466386583115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,3072,0.23598665661282012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,6144,0.18067910936143664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,4096,0.34961867332458496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,3584,0.38277244567871094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,2560,0.1850524478488498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,3072,0.10656088590621948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,3072,0.2705964512295193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,3584,0.30672889285617405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,2048,0.16237955623202854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,2560,0.10426755746205647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,3072,0.26872266663445366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,2560,0.21676799986097547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,2048,0.08147288693322076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,2048,0.17807377709282768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,1024,0.08527911371654934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,1536,0.0693368911743164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,1536,0.1423599984910753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,2048,0.1910711129506429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,768,0.06469421916537814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,1024,0.057686222924126514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,1024,0.10371466477711995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,1536,0.14892178111606175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,512,0.05204977922969394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,768,0.05445955528153313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,768,0.08258310953776042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,1024,0.11323289076487224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,256,0.034847110509872437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,512,0.0491484436723921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,512,0.06615733438067965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,768,0.09163644578721787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,128,0.026191999514897663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,256,0.05685600307252672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,512,0.07418400049209595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,64,0.022077333596017625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,1536,0.11828354994455974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,128,0.05267555514971415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,128,0.04462044437726339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,256,0.05493244528770447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,32,0.0233271105421914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,2560,0.22764266861809623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,64,0.051810665263069995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,32,0.05201955636342367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,256,0.04469155602984958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,128,0.04183822207980686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,65536,1.9157956441243489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,51200,1.3756560219658747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,51200,3.3921563890245228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,16384,1.2458035151163738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,16384,0.39602043893602157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,16384,1.2377750608656142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,65536,4.960394541422526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,12288,0.8253386815388998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,16384,1.2837502161661785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,51200,5.224415249294705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,51200,4.18339114718967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,12288,0.3230808840857612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,65536,5.383899688720703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,12288,0.8632675806681315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,10240,0.2693902121649848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,10240,0.7161751323276095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,8192,0.516632874806722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,10240,0.7781084378560384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,12288,0.9643413755628797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,8192,0.5742933485243056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,10240,0.7398266792297363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,8192,0.2773431142171224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,7168,0.4782266616821289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,8192,0.6060969034830729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,7168,0.18278133869171143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,6144,0.41933245129055446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,7168,0.5288577609592015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,6144,0.15969778431786433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,7168,0.5229511260986328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,5120,0.38370667563544375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,65536,8.052663167317709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,6144,0.6197964350382487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,5120,0.14021155569288465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,6144,0.45656622780693906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,4096,0.30871110492282444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,3584,0.21474933624267578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,5120,0.38040712144639754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,4096,0.38160888353983563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,3584,0.10537066724565293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,3072,0.1945128838221232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,3584,0.3693715466393365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,3072,0.09558933311038548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,2560,0.1716079976823595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,3584,0.2893608940972222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,3072,0.24175379011366102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,5120,0.3736097812652588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,2560,0.0861662228902181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,2560,0.19103733698527017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,2048,0.13138400183783636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,1536,0.0997155573632982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,2560,0.20805154906378853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,2048,0.07483377721574572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,2048,0.15720444255405003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,1536,0.06324888600243463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,1024,0.07490044169955783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,2048,0.170653329955207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,1536,0.12922044595082602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,4096,0.3199768861134847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,768,0.05132355623775058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,1024,0.05327199896176656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,1536,0.1342328919304742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,1024,0.09227466583251953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,512,0.04162399967511495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,3072,0.23080266846550834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,768,0.05031822125116984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,1024,0.10171644555197822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,768,0.07366489039527045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,256,0.02867822183503045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,512,0.045787556303872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,512,0.05836711327234904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,128,0.02054755555258857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,256,0.04168000155025058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,256,0.04972533384958903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,512,0.06636177831225924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,64,0.018226666582955253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,4096,0.11724089251624213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,32,0.01865600049495697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,128,0.04596889019012451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,256,0.04907733201980591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,64,0.04515199859937032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,128,0.03769777880774604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,32,0.04515555500984192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,768,0.0834382242626614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,128,0.041364444626702204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,65536,1.5504257414076064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,51200,1.2735510932074654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,16384,1.0259368684556749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,51200,2.9630311330159507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,16384,0.3497280014885797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,65536,3.9950425889756946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,16384,1.0550222396850586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,12288,0.677495108710395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,51200,4.292223188612196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,16384,1.1127582126193576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,12288,0.2935546769036187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,51200,3.6746258205837674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,12288,0.7815217971801758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,65536,4.737961663140191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,10240,0.49933958053588867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,10240,0.2324897713131375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,12288,0.8512835502624512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,65536,5.5414683024088545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,10240,0.6498835351732042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,8192,0.40312356419033474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,8192,0.1806711090935601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,7168,0.375091552734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,7168,0.17863733238644072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,8192,0.5392462412516276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,10240,0.7037750879923502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,6144,0.3380684322781033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,8192,0.5842773119608561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,7168,0.4760319921705458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,6144,0.14262844456566706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,5120,0.25807200537787545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,7168,0.4700337515936957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,6144,0.41814221276177305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,4096,0.2029404507742988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,5120,0.3830097781287299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,6144,0.4186560047997369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,4096,0.1128915548324585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,4096,0.3143884340922038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,5120,0.34271910455491805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,4096,0.28424355718824595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,3584,0.09561066495047675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,3072,0.15563378069135878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,3584,0.2289644347296821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,2560,0.1378364430533515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,3072,0.08612266514036392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,3584,0.2522835466596815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,3072,0.1985164483388265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,2048,0.11627999941507976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,2560,0.07741689019733004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,3072,0.21790666050381133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,5120,0.12358933024936253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,2560,0.16848444938659668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,1536,0.08395200305514866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,2048,0.06739466720157199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,2048,0.13767200046115452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,3584,0.19289510779910615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,1536,0.10866577757729425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,1536,0.05745333433151245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,1024,0.06087555487950643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,768,0.04820088876618279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,2048,0.15351822641160753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,1024,0.04887288808822632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,1536,0.12168267038133408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,1024,0.0802417794863383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,512,0.03216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,768,0.045907557010650635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,768,0.06408800019158258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,256,0.02241155505180359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,1024,0.09133155478371514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,512,0.041899555259280734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,512,0.05116800136036343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,768,0.07414133018917508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,128,0.018039999736679923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,256,0.04310933417744107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,64,0.017222222354676988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,512,0.05975288814968533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,128,0.037222223149405584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,32,0.01811555524667104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,128,0.0395671096112993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,256,0.04371733466784159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,2560,0.18444621562957764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,128,0.03315555718210008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,64,0.038466665479871966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,32,0.03805066810713874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,256,0.037964443365732826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,65536,1.2868533664279513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,51200,1.0148489210340712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,51200,2.2070045471191406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,16384,0.9205146365695529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,65536,2.8022516038682728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,16384,0.31283556090460884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,12288,0.5109333462185329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,16384,0.8904204898410373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,51200,3.145491494072808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,16384,0.9464844597710503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,51200,3.58601082695855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,12288,0.2309244473775228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,12288,0.6504835552639431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,10240,0.41797600852118594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,10240,0.20594844553205702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,8192,0.3377511236402724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,65536,4.034993913438585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,12288,0.6936133172776965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,10240,0.5450986756218804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,8192,0.15731910864512125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,7168,0.3236986796061198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,10240,0.5927119784884983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,8192,0.48646042082044816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,65536,5.177201588948567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,7168,0.14049777719709608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,6144,0.2506142192416721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,8192,0.4752426677280002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,5120,0.21833156214820015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,6144,0.14474933677249485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,7168,0.41249956025017637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,6144,0.34522223472595215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,4096,0.18042400148179796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,5120,0.12369688351949056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,6144,0.35066845681932235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,5120,0.41197776794433594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,3584,0.15241777896881104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,4096,0.09465333488252427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,4096,0.2216320037841797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,5120,0.2976435555352105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,3072,0.12969333595699734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,3584,0.09353599945704143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,4096,0.2454728815290663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,3584,0.19384800063239205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,2560,0.10903466410107082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,3072,0.0765119989713033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,3072,0.16878133349948457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,2048,0.0880168875058492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,3584,0.22580978605482313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,2560,0.07134222322040133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,3072,0.1898204485575358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,2560,0.1437271171145969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,1536,0.07162666983074613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,2048,0.06081510914696587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,2048,0.11699999703301324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,2560,0.15949155224694148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,1024,0.04609244399600559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,1536,0.05214488837454054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,1536,0.09241244528028701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,2048,0.13513422012329102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,768,0.03714133302370707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,1024,0.06816444132063124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,1024,0.04464888903829786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,1536,0.10604533221986558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,512,0.026535110341178045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,768,0.05510666635301378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,768,0.041336887412601046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,256,0.017416000366210938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,512,0.03528889020284017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,512,0.042846222718556724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,7168,0.39719290203518337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,512,0.05138044555981954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,256,0.03581066595183478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,128,0.014454222387737699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,128,0.028745777077145044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,256,0.03784711162249247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,64,0.013109332985348172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,128,0.03272533416748047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,32,0.013389333254761167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,128,0.028880000114440918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,64,0.031567109955681696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,32,0.031204445494545832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,1024,0.07943199740515815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,768,0.0638986693488227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,256,0.03190577692455716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,65536,1.0626684824625652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,65536,2.094214121500651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,51200,1.7619395785861547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,16384,0.6183066897922093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,51200,0.8772017690870496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,16384,0.285899559656779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,16384,0.7243999905056424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,12288,0.48551379309760195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,51200,2.6594835917154946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,12288,0.23006667031182182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,16384,0.8265804184807671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,12288,0.5344284375508627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,65536,3.5750855339898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,10240,0.3163093460930718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,10240,0.17217422856224907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,65536,3.433468500773112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,8192,0.26514045397440594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,12288,0.6009742418924967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,10240,0.47383822335137266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,8192,0.13294043805864122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,7168,0.2450186676449246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,8192,0.39567822880215114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,10240,0.5086755752563477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,7168,0.11858310964372422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,6144,0.2217644453048706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,7168,0.31433333290947807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,8192,0.4058853255377875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,6144,0.10426133208804661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,5120,0.18773867024315727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,6144,0.27484001053704155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,7168,0.35695555475023055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,5120,0.0905102226469252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,5120,0.22523999214172363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,51200,2.6632845136854386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,6144,0.3148275481330024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,3584,0.13306844234466553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,4096,0.07752711243099637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,4096,0.18314754962921143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,5120,0.26501867506239146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,3072,0.10903466410107082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,3584,0.07210222217771742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,3584,0.16069422827826604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,4096,0.21330755286746553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,2560,0.08616177903281318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,3072,0.06638399759928386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,3072,0.1393244398964776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,2048,0.0733226670159234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,3584,0.18885155518849692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,2560,0.06105066670311821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,1536,0.05567110909356011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,2560,0.1191742287741767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,2048,0.052770667605929904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,2048,0.09714666340086196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,2560,0.1416915522681342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,1024,0.03650310966703627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,1536,0.045999109745025635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,4096,0.15213067001766628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,2048,0.11767644352383083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,1536,0.07663822174072266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,768,0.029527111185921565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,1024,0.03698488738801744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,1536,0.09262577692667644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,768,0.03160888950030009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,1024,0.05637866920895047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,768,0.04565866788228353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,1024,0.06855911016464233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,256,0.014503111441930136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,768,0.05594578054216173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,512,0.035150223308139376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,512,0.02743822170628442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,128,0.011934222446547614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,256,0.024085332949956257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,256,0.028396444188223943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,128,0.026131555438041687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,3072,0.16658933957417807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,64,0.010841777755154504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,128,0.024051555328898962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,256,0.03130844566557143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,32,0.011229332950380115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,128,0.02405511173937056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,64,0.024703111913469102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,32,0.024346666203604803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,512,0.019863999552196927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,512,0.04401866594950358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,65536,0.8763830926683215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,65536,1.4853760401407878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,51200,1.1770897971259224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,16384,0.391396443049113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,51200,0.6812746789720324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,12288,0.28162044949001735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,16384,0.2280062304602729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,16384,0.6056835386488173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,51200,1.842864990234375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,65536,2.5145386589898004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,16384,0.6694826549953885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,12288,0.16342133945888943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,10240,0.23218221134609648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,12288,0.46413776609632706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,10240,0.14105689525604248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,10240,0.36240800221761066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,12288,0.5039653248257107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,65536,2.752690633138021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,51200,2.1542587280273438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,7168,0.1641599999533759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,10240,0.4147644572787815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,8192,0.280457788043552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,6144,0.13420355319976807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,7168,0.0965955588552687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,8192,0.33365776803758407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,7168,0.25703822241889107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,5120,0.11677955256568061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,6144,0.08534222178988987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,6144,0.2125840054617988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,7168,0.29345777299669057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,4096,0.08989155292510986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,5120,0.07440800136990018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,5120,0.18267377217610678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,8192,0.2020719978544447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,6144,0.2573920090993245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,3584,0.07785511016845703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,4096,0.14465155866410997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,4096,0.06527111265394422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,3072,0.0765777760081821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,3584,0.06002933449215359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,3584,0.1267297797732883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,4096,0.1735431088341607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,8192,0.10780621899498834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,2560,0.05889866749445597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,3072,0.11047910981708103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,3584,0.1576204432381524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,2048,0.046111110183927745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,2560,0.05137955480151706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,2560,0.09432178073459202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,3072,0.13749955760108099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,1536,0.03564088874393039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,2048,0.044770667950312294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,2048,0.07710311147901747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,2560,0.11731999450259739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,1024,0.0244835548930698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,1536,0.03696711195839776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,1536,0.060489780373043485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,768,0.01919822229279412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,1024,0.026884444885783728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,1536,0.07517955700556438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,5120,0.2150346702999539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,1024,0.055797331862979464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,768,0.03516088922818502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,768,0.024319999747806128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,3072,0.05595022439956665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,512,0.02187555531660716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,256,0.010692444112565784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,512,0.027238221632109746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,768,0.0458568897512224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,128,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,256,0.020816889074113634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,512,0.03498933381504483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,256,0.022111111217074927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,64,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,128,0.019709333777427673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,128,0.020592888196309406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,32,0.008215111162927415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,256,0.026304889056417678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,2048,0.09680977794859146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,64,0.01828977796766493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,128,0.02030044462945726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,32,0.01795288920402527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,1024,0.043960889180501304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,512,0.013083555632167392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,51200,0.888678232828776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,65536,1.3205991321139865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,65536,0.8430390887790256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,16384,0.34623111618889707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,51200,1.5424391428629558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,16384,0.5189208984375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,65536,2.062257766723633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,51200,0.7018986807929145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,12288,0.1956391069624159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,16384,0.1982684400346544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,12288,0.15628711382548013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,10240,0.164502223332723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,16384,0.6214257876078287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,12288,0.36582665973239475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,51200,1.9450577629937067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,10240,0.13883645004696316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,65536,2.5302897559271917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,8192,0.1380257738961114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,7168,0.12440621852874756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,12288,0.4778461986117893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,8192,0.10435822274949814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,8192,0.2429200013478597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,6144,0.11379999584621853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,10240,0.38223732842339414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,7168,0.09428622325261433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,7168,0.21219378047519258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,5120,0.08710400263468425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,8192,0.31178487671746147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,6144,0.08169155650668673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,6144,0.1830702225367228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,4096,0.06877866718504164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,7168,0.2767431206173367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,5120,0.07232711050245497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,5120,0.15376178423563638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,3584,0.06414310799704658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,4096,0.06335822078916761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,4096,0.1252959966659546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,5120,0.2008862230512831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,3072,0.051896890004475914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,3584,0.11026400327682495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,3584,0.059064891603257924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,4096,0.16246399614546034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,2560,0.04695644312434726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,3072,0.05400000015894572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,3584,0.1460124386681451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,3072,0.09564799732632107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,2048,0.037567999627855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,2560,0.0817048880789015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,2560,0.049010667535993785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,3072,0.12825866540273032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,2560,0.10859733157687718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,2048,0.043162667089038424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,2048,0.06647021902932061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,1024,0.01812533371978336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,1536,0.03353510962592231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,1536,0.052766223748524986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,2048,0.09059911304050022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,768,0.014852444330851236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,6144,0.2378844420115153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,1024,0.037631998459498085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,1536,0.06986044512854682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,10240,0.31110045644972056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,512,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,768,0.030656887425316706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,1024,0.051071110698911876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,256,0.008298666940795051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,512,0.021520000365045335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,512,0.023711111810472276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,512,0.03212799959712558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,256,0.020516445239384968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,128,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,256,0.01868977811601427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,1536,0.026919111609458923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,64,0.008372444245550368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,256,0.02316355539692773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,32,0.007522666619883643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,128,0.016521778371598985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,64,0.01515111161602868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,128,0.01812799937195248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,1024,0.02551377813021342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,32,0.0148044443792767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,768,0.023436443673239812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,768,0.04161689016554091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,65536,0.762663099500868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,51200,0.6134106847974989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,65536,0.6375173462761773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,128,0.02033866610791948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,16384,0.20466221703423393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,16384,0.18547022342681885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,51200,0.6086693339877659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,12288,0.15655467245313856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,16384,0.42080089781019425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,65536,1.6232817967732747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,51200,1.269336912367079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,10240,0.12371022171444362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,16384,0.531018680996365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,12288,0.3203973240322537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,10240,0.2544142140282525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,10240,0.11379733350541855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,8192,0.09132355451583862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,12288,0.4121093220180935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,65536,2.144806120130751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,51200,1.6776933670043945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,7168,0.08730933401319717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,10240,0.35036089685228133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,8192,0.08469777637057835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,8192,0.20662311712900797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,6144,0.07303822040557861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,7168,0.07708089219199286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,7168,0.18122488922542998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,5120,0.06440444125069512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,8192,0.2749697897169325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,6144,0.06788088877995808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,6144,0.155422223938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,12288,0.12792444229125977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,4096,0.04810755451520284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,7168,0.2422586547003852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,5120,0.059658666451772056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,6144,0.20947466956244576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,5120,0.1305475499894884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,3584,0.04718933171696133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,4096,0.05252444412973192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,3072,0.037575112448798284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,4096,0.10667555862002903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,5120,0.17710577117072213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,3584,0.04941511154174805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,2560,0.03224533465173509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,3584,0.0938639971945021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,3072,0.045216000742382474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,3072,0.08127111196517944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,2048,0.02418222195572323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,3584,0.12928088506062826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,2560,0.04126755396525065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,2560,0.06911022133297391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,1536,0.019574221637513902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,3072,0.11213510566287571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,2048,0.03538666831122504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,2048,0.05634311172697279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,1024,0.01369866645998425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,2560,0.09619822104771932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,1536,0.024661334024535284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,768,0.011128000087208219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,1536,0.04426577687263489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,2048,0.07996622059080336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,1024,0.021360889077186584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,512,0.008858666651778752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,1024,0.03183733423550924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,768,0.020279111133681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,768,0.025787555509143408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,256,0.006504888749784893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,1024,0.045067555374569364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,512,0.018590221802393597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,512,0.01994311147265964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,128,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,768,0.03587999939918518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,256,0.017907554904619854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,256,0.015465777781274585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,4096,0.14476977454291448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,64,0.005790222022268508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,32,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,128,0.01791288952032725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,128,0.013408888545301227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,256,0.021128000484572515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,64,0.012071111136012606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,128,0.015570667054918079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,32,0.0114720000161065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,65536,0.46513689888848203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,51200,0.3451964325375027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,1536,0.06107378005981445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,16384,0.10732977920108372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,65536,0.628604465060764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,512,0.027783112393485174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,51200,0.5057964324951172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,16384,0.32850755585564506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,65536,1.2839031219482422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,16384,0.15913511647118464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,12288,0.07825066645940144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,10240,0.06599377923541598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,12288,0.12587200270758733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,12288,0.24667376942104766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,16384,0.4696257909138997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,10240,0.22063110934363472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,8192,0.05611733595530192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,10240,0.11111199855804443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,65536,1.818991978963216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,12288,0.3498222298092312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,7168,0.048341333866119385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,8192,0.08277866575453016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,51200,1.441979514227973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,8192,0.17309421963161895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,6144,0.042344888051350914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,7168,0.14750399854448107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,8192,0.23714311917622885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,51200,1.0130062103271484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,5120,0.035637332333458796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,7168,0.2104382250044081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,6144,0.12825332747565374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,6144,0.06578310992982653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,4096,0.02975999977853563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,5120,0.05793333053588867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,5120,0.10787555244233872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,6144,0.18207022878858778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,3584,0.026524444421132404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,4096,0.08785688877105713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,4096,0.05046844482421875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,3584,0.07721422115961711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,3072,0.021687111920780603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,4096,0.12468177742428249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,10240,0.2931537893083361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,7168,0.07443644603093465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,3072,0.04338755541377597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,2560,0.01719111038578881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,3072,0.06700799862543742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,2560,0.03968177901373969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,3072,0.09824977980719672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,2048,0.013722666435771518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,2560,0.05652088589138455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,2560,0.08366933133867051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,1536,0.010980444649855295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,2048,0.045909331904517285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,5120,0.1530488861931695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,2048,0.0682720012134976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,1536,0.023564444647894964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,3584,0.047036445803112455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,1024,0.008433777425024245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,1536,0.03602666656176249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,768,0.0070088886552386815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,1024,0.02104088829623328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,1536,0.05271822214126587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,1024,0.025752888785468206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,3584,0.11158755090501572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,768,0.01999200052685208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,1024,0.036694222026401095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,512,0.016223111086421542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,768,0.03091288937462701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,512,0.018538667096032035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,256,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,256,0.017504889104101393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,2048,0.031066666046778362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,512,0.0240320000383589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,128,0.003891555385457145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,256,0.012099555797047086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,64,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,128,0.01756088932355245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,256,0.017439110411538016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,32,0.0038657776183552216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,128,0.00999022192425198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,64,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,128,0.013438222308953604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,32,0.008628444539176093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,65536,0.3337635464138455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,512,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,768,0.021006221572558086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,51200,0.263683557510376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,16384,0.08259377876917522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,65536,0.6341706911722819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,16384,0.3271511130862766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,51200,0.4923982090420193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,12288,0.06526577472686768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,16384,0.16239999400244817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,65536,1.162409782409668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,51200,0.9056790669759115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,16384,0.4335813257429335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,10240,0.05475199884838528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,12288,0.2355431185828315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,12288,0.12798133161332872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,8192,0.045936889118618436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,10240,0.11143555906083848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,10240,0.18964000542958578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,12288,0.32844887839423287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,7168,0.039396444956461586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,8192,0.08248177501890394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,8192,0.1550853384865655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,51200,1.3490284813774958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,6144,0.034747554196251765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,7168,0.07441955804824829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,7168,0.135517332288954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,5120,0.029317332638634577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,6144,0.0654346677992079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,7168,0.1940080059899224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,6144,0.11947555012173122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,4096,0.024280000064108107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,5120,0.09875466426213582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,5120,0.05786222219467163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,4096,0.05044355657365587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,3584,0.021592888567182753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,5120,0.14309155941009521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,4096,0.1162746614880032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,65536,1.6889609230889215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,3584,0.07031910949283175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,3072,0.017625777257813346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,3584,0.04759200082884895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,8192,0.22662221060858834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,3584,0.10357955429289077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,2560,0.010063110954231685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,3072,0.04412000046836006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,3072,0.06073244412740072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,3072,0.09098666906356812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,2560,0.03930044505331252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,6144,0.1699084440867106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,2048,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,2560,0.07706311013963488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,1536,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,4096,0.08014133241441515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,2048,0.0414613319767846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,2048,0.06195022000206841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,1536,0.03226577904489305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,1024,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,10240,0.2760008970896403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,1536,0.048039999273088246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,768,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,1024,0.02306844459639655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,1024,0.02128177715672387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,1024,0.03407377666897244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,768,0.0199306673473782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,512,0.004209777961174647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,768,0.01881866653760274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,2560,0.05117244521776835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,256,0.0035137778355015647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,512,0.018582221534517076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,768,0.02809422214825948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,512,0.014670222997665405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,128,0.0032133333798911837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,2048,0.02979644470744663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,256,0.01772444446881612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,256,0.010440888504187265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,64,0.002842666581273079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,128,0.017551110850440133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,128,0.008291555775536431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,32,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,1536,0.02369333306948344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,256,0.016022221909628976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,64,0.007254222200976477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,32,0.008271999657154083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,51200,0.24966666433546278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,65536,0.32703110906812877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,16384,0.08059200313356188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,12288,0.0626737806532118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,10240,0.05371288789643181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,8192,0.04493155413203769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,16384,0.29185332192314994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,12288,0.21681777636210123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,7168,0.03869688842031691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,6144,0.032906668053732976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,10240,0.1858666605419583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,512,0.021962665849261816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,5120,0.02877955635388692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,8192,0.1494737731085883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,4096,0.024754666619830664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,51200,0.8710622257656522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,3584,0.02070844504568312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,65536,1.1129030651516385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,7168,0.1299982203377618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,3072,0.01605955594115787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,2560,0.009130666653315226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,6144,0.11266399754418267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,2048,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,5120,0.09432266818152533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,1536,0.0075253330998950535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,1024,0.00572533326016532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,4096,0.0757697754436069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,768,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,3584,0.06647200054592557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,3072,0.05700977643330892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,2560,0.048362665706210665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,512,0.004229333251714706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,256,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,1536,0.030753778086768255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,128,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,1024,0.02211911148495144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,64,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,32,0.0027093332674768236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,768,0.017874666386180453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,512,0.013752889302041797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,256,0.00962933318482505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,128,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,64,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,32,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,16384,0.07821866538789538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,65536,0.35288532574971515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,12288,0.06044533517625597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,10240,0.050112889872656934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,51200,0.26323289341396755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,8192,0.044347554445266724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,16384,0.28720800081888836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,12288,0.2166879971822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,7168,0.03627644313706292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,10240,0.18249777952829996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,2048,0.03936533464325799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,8192,0.1486311091317071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,5120,0.027572444743580286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,51200,0.8697039816114637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,4096,0.02256800068749322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,7168,0.12979200151231554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,65536,1.1124987072414823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,3584,0.019581332802772522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,3072,0.01516266663869222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,6144,0.1125768952899509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,2560,0.00792622235086229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,5120,0.09420977698432074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,2048,0.006678222368160884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,4096,0.07583555248048571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,3584,0.06974577903747559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,1536,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,1024,0.004567999806669024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,3072,0.0609991086853875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,768,0.003930666794379552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,2560,0.050844444168938525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,512,0.0032017777363459268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,2048,0.04144444399409824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,1536,0.03216622273127238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,256,0.002624888800912433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,128,0.0022942223068740633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,1024,0.022791999909612868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,64,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,768,0.01854222185081906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,512,0.014145778285132514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,256,0.009748444788985783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,128,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,64,0.006684444430801604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,32,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,32,0.002159111201763153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,6144,0.031643556223975286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,16384,11.58161841498481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,51200,13.612775166829428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,16384,12.355379740397135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,16384,4.022624121771918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,128,0.01274311128589842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,16384,18.44863552517361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,12288,2.7843316396077475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,12288,8.827400207519531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,10240,6.660918341742621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,12288,11.899735344780815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,12288,9.66353013780382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,10240,8.248622470431858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,10240,2.3351412879096136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,8192,1.7254471249050563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,10240,7.519758436414931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,8192,5.280223846435547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,51200,33.36909654405382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,8192,6.010796440972222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,8192,5.516674465603298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,7168,1.4942053688897028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,7168,5.1120889451768665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,7168,4.940382215711805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,6144,1.3482693566216364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,6144,4.230188581678602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,7168,4.654395633273654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,6144,4.033752865261502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,5120,1.079399956597222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,6144,3.8773778279622397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,5120,3.5384203592936196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,4096,2.559268527560764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,5120,3.518037372165256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,4096,0.8759919802347819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,5120,3.2805724673800998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,4096,2.7457023196750217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,4096,2.6959039900037975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,3584,0.82150665918986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,3584,2.2254596286349826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,3584,2.3694337209065757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,3072,0.67431730694241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,3584,2.332888921101888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,3072,2.090097851223416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,3072,2.0244568718804254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,3072,2.0022542741563583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,2560,0.6414711210462782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,2560,1.8860187530517578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,2048,1.4301724963717992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,2560,1.7207697762383356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,2048,0.5963288942972819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,2048,1.368527094523112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,1536,1.0516622331407335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,2048,1.3980595270792644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,1536,0.4104026688469781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,1536,1.0282711452907987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,2560,1.682760026719835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,1024,0.7471493085225424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,1536,1.0488061904907227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,1024,0.31405509842766655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,1024,0.7886160214742025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,1024,0.7363084687127007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,768,0.2711564434899224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,512,0.34784711731804746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,768,0.9075466791788737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,512,0.2597271071539985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,512,0.5187457932366265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,256,0.2571546766493055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,512,0.4690382215711806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,768,0.5055128733317057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,128,0.21078844865163168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,256,0.22806755701700845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,256,0.46812089284261066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,256,0.3605582184261746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,64,0.19957599374983045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,128,0.43385688463846844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,32,0.19245689445071748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,768,0.5968853102789985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,128,0.2753386762407091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,64,0.4492799970838759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,32,0.4669182035658095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,128,0.21422133180830213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,51200,63.5065171983507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,65536,13.35001966688368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,51200,25.556747436523438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,51200,10.547471788194445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,65536,33.248209635416664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,16384,8.240262349446615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,16384,3.061642752753364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,16384,14.619210137261284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,51200,46.27781846788195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,16384,10.347403632269964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,12288,6.6857189602322045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,12288,2.2172889709472656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,10240,5.221959008110894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,12288,8.931334601508246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,51200,49.67064412434896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,10240,1.8643165164523656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,12288,7.312764485677083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,51200,37.09452650282118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,10240,6.535337660047744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,8192,1.3290231492784288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,8192,4.947998046875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,10240,5.960588243272569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,8192,4.621170467800565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,7168,3.561179690890842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,8192,4.379773457845052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,7168,1.1881244447496202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,7168,3.852635701497396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,6144,1.0921999613444011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,6144,3.0610889858669705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,6144,3.1187572479248047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,7168,3.6222381591796875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,5120,0.8683857917785645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,6144,3.0533154805501304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,5120,2.501385794745551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,5120,2.666735119289822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,4096,0.7161555820041232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,5120,2.561258739895291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,4096,2.000067604912652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,4096,2.2149225870768228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,3584,0.6121164427863227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,4096,2.013779534233941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,3584,2.0136595831976996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,3584,1.9966212378607855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,3584,1.8125919765896266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,3072,0.6309262381659614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,3072,1.8883840772840712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,3072,1.6364551120334203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,3072,1.5683555603027344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,2560,0.6206462118360732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,2560,1.2628213034735787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,2560,1.3576427035861547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,2048,1.0046524471706815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,2560,1.3393706215752497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,2048,0.43292800585428876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,2048,1.1003804736667209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,1536,0.8209484418233236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,1536,0.31799377335442436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,1024,0.5304284625583225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,1536,0.8425306744045682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,2048,1.0366053051418729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,1536,0.8061688741048177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,1024,0.3209751182132297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,768,0.4427288903130426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,1024,0.951264911227756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,768,0.2566506597730848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,1024,0.5740684403313531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,768,0.5131680170694987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,512,0.41318133142259383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,65536,69.64051649305556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,256,0.20915022161271837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,512,0.4017768965827094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,256,0.1843795511457655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,128,0.17682133780585396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,512,0.37525778346591526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,128,0.17074755827585855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,256,0.5558800167507595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,256,0.28496355480617946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,128,0.3479119936625163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,64,0.1498639980951945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,32,0.14475556214650473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,128,0.21694489320119223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,768,0.47382132212320965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,512,0.19746132691701254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,64,0.35843200153774685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,32,0.3692951202392578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,65536,5.030313703748915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,51200,8.02830081515842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,65536,10.140897962782118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,51200,4.131578657362196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,16384,2.6684159172905813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,16384,1.1009368896484375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,51200,11.737345377604166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,16384,3.316385693020291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,12288,2.032385720147027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,16384,4.723291609022352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,12288,0.8568942281934949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,12288,2.900100496080187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,51200,16.00018310546875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,65536,15.097312079535591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,12288,2.344672097100152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,10240,0.7244506412082248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,10240,1.9918418460422094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,8192,1.4794044494628906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,10240,1.9587785932752821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,8192,1.3559172948201497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,8192,0.5800657802157932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,7168,0.45268620385064023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,8192,1.4007590611775715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,10240,1.9408914777967665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,7168,1.1557084189520943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,7168,1.2466772927178276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,6144,1.077606201171875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,6144,0.347126219007704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,6144,1.023930655585395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,7168,1.2362808651394315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,6144,0.966477288140191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,5120,0.8293040063646105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,5120,0.3021022213829888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,4096,0.6649066607157389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,4096,0.24320443471272787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,5120,0.9629430770874023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,4096,0.6842568715413412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,5120,0.7657599978976779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,3584,0.21303289466434053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,3584,0.636664867401123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,4096,0.6230435371398926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,3584,0.5798924234178331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,3072,0.49086311128404403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,65536,22.55094061957465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,3072,0.18836711512671578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,2560,0.39348000950283474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,3072,0.5186995400322808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,2560,0.16622311539120144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,3072,0.5497733222113715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,2560,0.5205075475904677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,2560,0.39303021960788304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,2048,0.34908355606926816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,2048,0.13729333877563477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,1536,0.24995909796820748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,2048,0.330367989010281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,1024,0.17086844974093965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,1536,0.11199022663964166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,3584,0.5349022017584907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,1536,0.26895199881659615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,1024,0.08728710810343425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,768,0.14217421743604872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,1024,0.20283110936482748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,1536,0.25638045205010307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,512,0.10187377532323201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,768,0.07611111137602064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,1024,0.19297067324320474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,2048,0.34127023484971786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,256,0.07237688700358073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,512,0.06850044594870673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,512,0.1340577734841241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,768,0.15947911474439833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,128,0.055825776524013944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,256,0.06218577755822075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,512,0.12924355930752224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,256,0.12036889129214817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,64,0.046741333272722035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,128,0.06299377812279595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,128,0.11349066098531087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,32,0.047988444566726685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,256,0.10086933109495376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,128,0.076401776737637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,64,0.11712355083889431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,32,0.12239200539059109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,768,0.15994755427042642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,65536,47.364291720920136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,65536,3.7864646911621094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,51200,3.107729805840386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,16384,2.407513724433051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,51200,6.7593260871039496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,16384,0.870704862806532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,65536,9.019226921929254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,16384,3.539848963419596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,51200,8.94972652859158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,16384,2.5116719139946833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,12288,1.8542416890462239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,12288,0.6131893263922797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,10240,1.4211111068725586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,12288,2.2516985999213324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,12288,1.8618755340576172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,10240,1.5577155219184027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,10240,0.5169155332777235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,8192,1.0388675265842013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,10240,1.3274933497111003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,8192,0.3778008884853787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,65536,11.75970204671224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,51200,12.335505167643229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,8192,1.0617129007975261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,8192,1.2627422544691298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,7168,0.990015983581543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,7168,0.3978462219238281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,6144,0.7039404445224338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,7168,1.0209555096096463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,7168,0.8643049134148492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,6144,0.3208479881286621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,5120,0.6455635494656032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,6144,0.7912124527825249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,5120,0.2208791176478068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,5120,0.6312488979763454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,6144,0.6977893511454264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,5120,0.5791546503702799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,4096,0.5576835738288032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,4096,0.1825662189059787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,4096,0.5064764552646214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,3584,0.42787732018364805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,3584,0.1622355514102512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,4096,0.47249152925279403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,3072,0.3510062164730496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,3584,0.45380708906385636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,3072,0.14626044697231716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,3584,0.43498844570583767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,3072,0.48619556427001953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,3072,0.36320800251430935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,2048,0.2466071181827121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,2048,0.10767910877863567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,2048,0.2670737902323405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,1536,0.18719289037916395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,65536,16.445402357313366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,2048,0.2556240028805203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,1536,0.20675467120276558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,1024,0.14119288656446669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,1536,0.08674133486217922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,2560,0.33450667063395184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,1024,0.06855999761157565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,768,0.11325244108835857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,1024,0.1556968953874376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,2560,0.12535999880896673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,512,0.08137777778837416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,1024,0.15074933899773493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,768,0.06081333425309923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,768,0.1255226665072971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,2560,0.30629510349697536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,256,0.05477689041031731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,512,0.05548000004556444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,768,0.12543377611372206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,512,0.10308177603615655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,128,0.03856533434655931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,256,0.05095466640260485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,512,0.10117689106199478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,256,0.09233778052859837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,64,0.03575911124547323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,128,0.05073688758744133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,128,0.08695555395550197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,32,0.03569599986076355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,256,0.07882489098442926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,128,0.06007644202974108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,64,0.08931555350621541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,1536,0.19905332724253336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,32,0.09330666727489895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,2560,0.3288195663028293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,51200,2.491593678792318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,65536,3.2449510362413196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,16384,1.835998323228624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,51200,5.216116587320964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,16384,0.6521626578436958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,65536,6.864806281195746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,16384,3.0304489135742188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,16384,1.9800239139133031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,12288,1.3995964262220595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,12288,0.514476458231608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,51200,7.61307356092665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,12288,1.4654738108317058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,12288,1.9295048183865016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,10240,1.0625982284545898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,10240,0.4266808827718099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,8192,0.8444942368401421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,10240,1.1998817655775282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,8192,0.9377617306179471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,51200,10.14896731906467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,8192,0.8515173594156901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,7168,0.8172222243414985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,10240,1.2633288701375325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,7168,0.2797795666588677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,7168,0.7524391280280219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,65536,9.48291015625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,6144,0.22450576888190377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,6144,0.7001075744628906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,7168,0.7002044253879123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,6144,0.6360541979471842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,5120,0.18566132916344535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,5120,0.5253840022616917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,6144,0.5998249053955078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,5120,0.5463920169406468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,4096,0.1576391061147054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,5120,0.5180675718519423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,4096,0.43339912096659344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,3584,0.3410702281528049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,8192,0.29735999637179905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,4096,0.4123724566565619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,3584,0.14139200581444636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,3584,0.3863617844051785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,3072,0.3234684467315674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,3072,0.3191537857055664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,3072,0.12679021888309056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,3584,0.36231289969550234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,2560,0.269815100563897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,4096,0.43064265780978733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,2048,0.23449688487582734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,3072,0.31083112292819554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,2560,0.2715208795335558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,2048,0.09280266364415486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,1536,0.13759377267625597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,2560,0.2640257676442464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,2048,0.22215822007921007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,1024,0.11281511518690322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,1536,0.07561422056621976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,1536,0.17490932676527235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,2048,0.3072506586710612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,65536,14.294521755642362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,768,0.08304266797171699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,1536,0.1708737744225396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,1024,0.05945510996712578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,1024,0.1323537826538086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,512,0.0657822224828932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,768,0.05972444348865085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,768,0.10466666354073419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,1024,0.13055377536349827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,2560,0.11066489087210761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,512,0.08769155873192681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,768,0.11113156212700738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,128,0.03651288814014859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,256,0.044695999887254506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,512,0.08790666527218288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,256,0.07810311184989081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,64,0.030470222234725952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,32,0.032440887557135686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,128,0.044933332337273486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,256,0.06759822368621826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,128,0.0737537807888455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,128,0.05159822106361389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,64,0.07543288999133639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,32,0.07890666855706109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,256,0.04606222112973531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,512,0.04889422323968676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,51200,1.97850587632921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,65536,2.6046454111735025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,16384,1.4900222354465062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,51200,4.481089697943793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,65536,5.21040513780382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,16384,0.4998817973666721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,16384,2.3792879316541886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,12288,1.1486426459418404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,16384,1.6719946331448023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,12288,0.4538755416870117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,12288,1.4936951531304254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,51200,5.932248857286241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,12288,1.1371741824679904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,10240,0.9755964279174805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,10240,0.35273154576619464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,65536,7.4999847412109375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,8192,0.7509466807047526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,8192,0.3008435567220052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,10240,1.155893325805664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,10240,0.9691164228651258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,51200,8.289666069878473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,7168,0.6379084587097168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,7168,0.234040896097819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,8192,1.0340471267700195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,8192,0.7134390936957465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,6144,0.5063333511352539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,6144,0.1784586641523573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,7168,0.5835928916931152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,7168,0.793184863196479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,5120,0.3925413290659587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,6144,0.5731546613905165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,5120,0.153912001185947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,4096,0.28581778208414715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,6144,0.5028586917453343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,5120,0.4854026900397406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,4096,0.12876444392734104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,3584,0.2870453463660346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,4096,0.36229422357347274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,5120,0.4052533308664958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,3584,0.11679822868771023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,3072,0.2536444399091932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,4096,0.3317146566179064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,3584,0.31317244635687935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,3072,0.10362578100628322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,3072,0.26153598891364205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,2560,0.21907822291056314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,3584,0.2940151161617703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,2560,0.09070133500629002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,3072,0.2521057658725315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,2560,0.22208800580766466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,65536,10.740966796875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,1536,0.13219822777642146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,2048,0.0765226682027181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,2048,0.22016089492373994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,1024,0.09528444210688274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,1536,0.06195555792914497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,2048,0.18071377277374268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,1536,0.1432808902528551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,768,0.07696444458431669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,1024,0.11266577243804932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,1536,0.14350667264726427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,1024,0.051382223765055336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,512,0.054216000768873424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,768,0.04661422305636936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,1024,0.10820888810687596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,768,0.08636266655392116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,256,0.03722044494416978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,512,0.042018665207756885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,512,0.07214044200049506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,768,0.09029599693086411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,2048,0.1782133314344618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,128,0.03013422091801961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,256,0.03925688730345832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,256,0.06418577829996745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,64,0.024717332588301763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,2560,0.21465599536895752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,128,0.03916088740030924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,32,0.025196444657113817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,128,0.06007644202974108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,256,0.05624177720811632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,128,0.04372088776694404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,64,0.061612447102864586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,32,0.06443555487526788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,512,0.07376711236106025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,65536,2.189200931125217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,51200,1.652901331583659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,51200,4.018726348876953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,16384,1.4767395655314128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,16384,0.46880801518758136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,65536,5.621989356146918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,16384,1.9255875481499567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,12288,1.0913217332628038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,51200,4.82804701063368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,12288,0.3247671127319336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,51200,6.67398664686415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,12288,1.0255582597520616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,65536,6.784024132622613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,16384,1.6198470857408311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,10240,0.2938320106930203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,10240,0.9702026579115125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,10240,0.8173084259033203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,8192,0.2324915462070041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,10240,0.8737413088480631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,12288,1.0795858171251085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,8192,0.8569502300686307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,8192,0.6050408681233724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,7168,0.57679197523329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,7168,0.18533778190612793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,8192,0.628894223107232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,6144,0.16106577714284262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,7168,0.5506470998128256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,6144,0.5919546551174587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,6144,0.45337777667575413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,5120,0.3881262143452962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,7168,0.6901911099751791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,5120,0.15915200445387098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,4096,0.3263048860761854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,6144,0.451524469587538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,5120,0.37952444288465714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,4096,0.11837155289120144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,3584,0.2833191023932563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,5120,0.36872177653842503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,65536,9.560869852701822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,3584,0.2728399965498182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,4096,0.3214915593465169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,3072,0.26068001323276097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,3584,0.26098044713338214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,3072,0.09188710980945164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,3072,0.2413733270433214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,2048,0.14277599917517766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,2560,0.0806257790989346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,3072,0.2295600043402778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,2560,0.24703555636935762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,1536,0.12574756145477295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,2048,0.06877599822150336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,2560,0.19438489278157553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,2048,0.16611999935574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,1024,0.07543822129567464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,1536,0.05605510870615641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,4096,0.3538826571570502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,2048,0.16257422500186497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,3584,0.10616888602574666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,1536,0.13525956206851536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,768,0.07155377997292413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,1024,0.04650400082270304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,1024,0.09593333138359918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,1536,0.12930133607652453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,768,0.0430026650428772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,768,0.07720888985527886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,2560,0.1804880036248101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,1024,0.09833955764770508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,256,0.03847555650605096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,512,0.03920977645450168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,768,0.08300799793667264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,128,0.0266311119000117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,512,0.06186133623123169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,256,0.03599822190072801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,256,0.054550220568974815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,512,0.06650311417049833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,64,0.02202133337656657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,32,0.023572444915771484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,128,0.03622755408287048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,256,0.051496889856126576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,128,0.039339555634392634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,64,0.051338666015201144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,32,0.05237866772545708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,512,0.05388977792527941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,128,0.051240887906816274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,65536,1.7991243998209636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,51200,3.0957315233018665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,51200,1.3490577273898656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,16384,1.2693920135498047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,65536,4.087627410888672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,16384,0.3516684373219808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,12288,0.8228169017367892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,16384,1.4565635257297094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,16384,1.224130630493164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,65536,5.377082824707031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,51200,4.136758168538411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,10240,0.6755608982510037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,51200,5.631503211127387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,12288,0.9195458094278971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,10240,0.2642844518025716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,12288,0.8960630628797742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,8192,0.5302666558159722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,8192,0.18306044737497965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,10240,0.978358268737793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,8192,0.7017751269870334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,10240,0.8326329125298394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,65536,7.466396755642361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,7168,0.5637377632988824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,7168,0.16033511691623265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,8192,0.5431831147935655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,6144,0.40189067522684735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,7168,0.4569919904073079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,6144,0.14324533939361572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,5120,0.3058319886525472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,12288,0.26191287570529515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,7168,0.45041688283284503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,5120,0.12325599458482529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,6144,0.4334968990749783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,4096,0.2559200127919515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,5120,0.42355023490058047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,4096,0.10252710845735337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,5120,0.32018311818440753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,4096,0.26556534237331814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,3584,0.09297066926956177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,3072,0.18290932973225912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,4096,0.27145777808295357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,3584,0.3222240077124702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,2560,0.15692978435092503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,3072,0.08274577723609076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,3584,0.2391902340783013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,2048,0.1341822279824151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,2560,0.0717537800470988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,6144,0.4131048785315619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,2560,0.1993511120478312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,3072,0.20038310686747232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,1536,0.1057422227329678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,2048,0.06046577956941393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,2560,0.17276444700029162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,2048,0.14416444301605225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,1024,0.07377777496973674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,3584,0.2235626644558377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,1536,0.11273333761427139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,2048,0.14539999432033962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,1024,0.04269955555597941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,1536,0.1156266662809584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,1024,0.08768444591098362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,768,0.03889333208401998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,3072,0.20266932911343047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,512,0.04169333312246534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,768,0.07345155874888103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,512,0.035755554835001625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,256,0.028235554695129395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,512,0.05454933312204149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,512,0.06004177861743503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,256,0.047813332743114896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,1536,0.05044710967275831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,256,0.04594399862819248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,768,0.056629333231184215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,64,0.01827111177974277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,128,0.044878221220440336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,128,0.03286222285694546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,32,0.018851555056042142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,1024,0.0840559999148051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,768,0.06821244292789035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,128,0.03525866733656989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,64,0.044575111733542554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,32,0.04524711105558607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,256,0.032687998480266996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,128,0.02077866593996684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,65536,1.3933804829915364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,51200,1.085982216729058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,16384,1.097428427802192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,51200,2.8629271189371743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,65536,3.2762328253852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,16384,1.0734969245062935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,16384,0.29982932408650714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,12288,0.7213999960157605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,51200,4.238822089301215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,16384,1.0773448944091797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,12288,0.7109795676337348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,51200,3.6322063869900174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,12288,0.22684889369540742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,10240,0.5499288770887587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,10240,0.19387555122375488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,12288,0.736462222205268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,65536,4.808587392171224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,10240,0.5706702338324653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,8192,0.43978222211201984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,8192,0.19329867098066542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,7168,0.3959706624348958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,65536,6.089539421929254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,10240,0.6420515378316244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,8192,0.5005804697672526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,7168,0.17548978328704834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,6144,0.12290399604373509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,6144,0.3573911190032959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,8192,0.5553670989142524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,7168,0.5542151133219401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,5120,0.27070487870110405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,6144,0.37136001057094997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,7168,0.39512356122334796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,5120,0.10479288631015354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,4096,0.22144888506995308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,5120,0.28692176606920033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,4096,0.08915466732449001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,3584,0.1911777787738376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,4096,0.24327378802829316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,3072,0.15166310469309488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,5120,0.33137689696417916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,3584,0.08064888583289252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,3584,0.21783733367919922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,4096,0.2320880095163981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,2560,0.13156710730658636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,3072,0.08078577783372667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,3584,0.20248622364468047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,3072,0.1782604455947876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,2048,0.11126844088236491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,2560,0.0631493330001831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,2560,0.15126755502488878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,1536,0.09025244580374824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,3072,0.1819333367877536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,2048,0.057505779796176486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,2560,0.15284177992078993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,2048,0.12290844652387832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,1536,0.04576711191071404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,1536,0.09745333592096965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,2048,0.1283386680814955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,768,0.04118488894568549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,6144,0.35472978485955137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,1024,0.03856888744566176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,512,0.031146665414174397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,1024,0.07281333208084106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,1536,0.10119199752807617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,768,0.035207112630208336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,256,0.022870222727457683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,1024,0.0779671139187283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,512,0.031834665271970965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,768,0.05924089087380303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,128,0.018146667215559218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,512,0.047016001409954496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,768,0.06433333290947808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,256,0.029376887612872656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,512,0.05337599913279215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,64,0.01733688844574822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,32,0.01850666602452596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,128,0.02937600016593933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,128,0.03837955660290188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,256,0.0410151117377811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,128,0.030722667773564655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,64,0.038091556893454656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,32,0.038894222842322454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,1024,0.05716088745329115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,256,0.0407244430647956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,65536,1.112213346693251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,51200,0.8575475480821398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,51200,2.1877386305067277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,16384,0.8538382318284777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,16384,0.26467378934224445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,65536,3.038295957777235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,51200,2.6105333964029946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,16384,0.795430236392551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,12288,0.5549946890936958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,16384,0.7837324672275119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,12288,0.1943288909064399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,51200,3.5353448655870228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,12288,0.5820479922824436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,65536,3.3682738410102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,10240,0.44418221049838597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,10240,0.16591821776496038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,12288,0.5695093472798666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,8192,0.3647884527842204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,8192,0.1309057739045885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,10240,0.5836462444729275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,65536,4.525143941243489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,8192,0.4018293221791585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,7168,0.2859120104047987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,10240,0.46059820387098527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,7168,0.12102933724721272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,6144,0.24896534283955893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,7168,0.3715875678592258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,8192,0.49547020594278973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,5120,0.20816355281405977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,6144,0.10354310936397976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,7168,0.3225439919365777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,4096,0.16238044367896184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,6144,0.28146311971876353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,5120,0.10263555579715306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,5120,0.241742213567098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,3584,0.1388337744606866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,4096,0.0756711098882887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,5120,0.27018578847249347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,4096,0.1889422204759386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,3072,0.12713778018951416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,3584,0.1663502189848158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,4096,0.19409333335028756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,2560,0.10127821895811294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,3072,0.14511111047532824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,3072,0.06053421894709269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,3584,0.17196267180972627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,2048,0.08288800054126315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,2560,0.05359466539488899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,3072,0.1520586676067776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,2560,0.12498222457038032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,6144,0.3651217884487576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,1536,0.0643048882484436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,2048,0.04677066538068983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,2560,0.1285359991921319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,2048,0.10791911019219293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,1024,0.049419555399152965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,1536,0.03962488969167074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,2048,0.10921244488822089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,1536,0.08132711384031507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,768,0.036708444356918335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,1024,0.06111911270353529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,1024,0.033511112133661904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,512,0.026856889327367146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,1536,0.08739378054936726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,3584,0.06816266642676459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,768,0.030180444320042927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,256,0.017669333351982962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,768,0.04936266607708401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,1024,0.06658044126298693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,128,0.014133334159851074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,512,0.03877955675125122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,768,0.05592799848980374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,256,0.02443733314673106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,64,0.012890666723251343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,256,0.03340977761480544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,32,0.013292444248994192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,512,0.04560533497068617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,128,0.03135111265712314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,256,0.03444622291458978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,64,0.030883553955290053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,128,0.02678577767478095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,32,0.031116445859273274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,512,0.026943999860021803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,128,0.024343111448817786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,65536,0.9844693077935113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,51200,1.9537902408176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,51200,0.7762311299641927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,16384,0.7950631247626411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,16384,0.2473564412858751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,65536,2.7973448435465493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,16384,0.6868160035875109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,51200,2.9370399051242404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,12288,0.5070035722520616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,16384,0.7010711034138998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,12288,0.17311999532911512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,12288,0.5472879939609104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,65536,3.109183205498589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,51200,2.3886612786187067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,65536,3.7539261711968313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,10240,0.14923911624484593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,8192,0.3057626618279351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,12288,0.5061733457777235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,10240,0.43176711930169004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,8192,0.12010844548543294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,7168,0.28246577580769855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,10240,0.42778934372795957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,8192,0.3545075522528754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,7168,0.10752444134818183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,8192,0.34896267784966367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,6144,0.24776710404290092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,7168,0.31277510854932994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,6144,0.09413422478569879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,7168,0.3041893376244439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,6144,0.25888800621032715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,5120,0.08141777912775676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,4096,0.15662133693695068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,10240,0.4002817736731635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,6144,0.2624044418334961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,5120,0.24086578687032065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,3584,0.13480977217356363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,4096,0.07628889216317071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,4096,0.17019555303785536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,3584,0.149308443069458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,3584,0.06617777877383761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,4096,0.18223999606238472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,3584,0.16304532686869302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,3072,0.05616355604595608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,2560,0.09422044621573554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,3072,0.13030844264560276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,3072,0.14637688795725504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,2560,0.050006220738093056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,2560,0.11243377791510688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,2048,0.07817511426077949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,2560,0.12232266532050239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,2048,0.04270755582385593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,1536,0.06832444667816162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,5120,0.22092621856265596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,2048,0.10312444633907741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,3072,0.12351377805074056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,1536,0.035972442891862653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,1536,0.07327999671300252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,1024,0.043839110268486865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,768,0.034556445148256086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,1024,0.02938844429122077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,1536,0.08136889007356432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,512,0.02560088866286808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,1024,0.05481333202785916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,768,0.044738666878806226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,768,0.027248889207839966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,256,0.01749955614407857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,5120,0.2008648845884535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,768,0.05255022313859728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,512,0.024030221833123103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,512,0.03480888737572564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,128,0.01445777714252472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,256,0.030023111237419978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,512,0.04268444577852885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,64,0.016237333416938782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,128,0.028147555059856836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,2048,0.09279200103547837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,128,0.02183466653029124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,32,0.016951999730534024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,256,0.03136533498764038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,128,0.02401688860522376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,64,0.027493332823117573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,32,0.027669333749347266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,1024,0.06286133660210504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,256,0.021493333909246657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,65536,0.8938692940606011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,16384,0.6546728875901964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,51200,0.6919475661383735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,51200,1.8237476348876953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,65536,2.175984912448459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,16384,0.2234720124138726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,12288,0.4169617758856879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,16384,0.7308088938395182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,16384,0.6315626568264431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,12288,0.46832799911499023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,51200,2.081922743055556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,51200,2.9145156012641062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,12288,0.15876000457339817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,65536,3.2648436228434243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,10240,0.3383644421895345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,10240,0.14344888263278538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,12288,0.4767288631863064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,8192,0.26443910598754883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,10240,0.36969778272840714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,8192,0.10751822259691025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,7168,0.2351048787434896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,8192,0.3586222330729167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,10240,0.3895706600613064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,7168,0.0961502194404602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,6144,0.19196177853478325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,7168,0.28913866149054634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,8192,0.35982134607103133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,5120,0.1691919962565104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,6144,0.22678311665852866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,7168,0.2710808912913005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,65536,2.7312488555908203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,4096,0.14289244016011557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,5120,0.1903724405500624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,6144,0.23747645484076607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,3584,0.12034311559465195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,4096,0.06220799684524536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,4096,0.1494657728407118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,5120,0.20395733250512016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,3072,0.09861422247356838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,4096,0.16297600004408094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,3072,0.0507226652569241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,3584,0.14663822121090359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,2560,0.08116177717844646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,6144,0.0848311119609409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,3072,0.1276115576426188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,2048,0.07299288776185778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,2560,0.0999786655108134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,2560,0.04554399847984314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,5120,0.08141155375374688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,1536,0.05032889048258463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,2560,0.10919021897845799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,2048,0.03800355394681295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,2048,0.08255555232365926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,2048,0.09336000018649632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,1536,0.03190222382545471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,3584,0.056757330894470215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,1024,0.03465955456097921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,3584,0.1331066687901815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,768,0.026559111144807603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,1024,0.02574133376280467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,3072,0.11692444483439128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,1024,0.048398223188188344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,512,0.01940266622437371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,768,0.039713776773876615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,768,0.023131555981106226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,1024,0.05609244770473904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,256,0.01421511173248291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,512,0.02085688875781165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,768,0.04730844497680664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,512,0.031179553932613794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,128,0.01202933324707879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,256,0.026392888691690233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,256,0.01964977714750502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,512,0.038634667793909706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,64,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,128,0.0196426659822464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,32,0.011235555840863122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,256,0.02910755409134759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,1536,0.06526133086946276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,64,0.02437244521247016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,128,0.021948445174429152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,1536,0.07321066988839044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,32,0.02440533373090956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,65536,0.7767297956678602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,128,0.02478755513827006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,16384,0.4445750978257921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,51200,1.5523999532063801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,65536,1.9104071723090277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,51200,0.6271760198805068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,16384,0.2267235649956597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,51200,2.0343769921196833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,12288,0.3337599966261122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,16384,0.5352097617255317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,65536,2.365655051337348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,12288,0.16066133975982666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,10240,0.28196801079644096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,51200,1.834457821316189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,12288,0.457764466603597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,10240,0.11889244450463189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,8192,0.22423733605278862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,12288,0.41097601254781085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,10240,0.3428506586286757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,8192,0.09872355726030137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,7168,0.20178043842315674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,8192,0.26123910480075413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,10240,0.3660542170206706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,6144,0.1707893345091078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,7168,0.08584177494049072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,8192,0.2791351212395562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,7168,0.23080621825324166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,6144,0.07555555635028414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,5120,0.1443342235353258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,6144,0.20144711600409615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,7168,0.24758489926656088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,65536,2.7608044942220054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,16384,0.5862675772772895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,4096,0.11763111750284831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,5120,0.06619377930959065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,5120,0.16591466797722712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,3584,0.10713599787818061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,6144,0.21494932969411215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,4096,0.05505599909358554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,4096,0.13223466608259413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,3072,0.0893262227376302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,3584,0.05054755674468147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,5120,0.17795556121402314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,3584,0.1165742211871677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,4096,0.14655555619133845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,2560,0.07310577895906237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,3072,0.045759111642837524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,3072,0.10185511244667901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,2048,0.061108443472120494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,3584,0.13068889247046575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,2560,0.03939911060863071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,1536,0.04453777935769823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,2048,0.03381777803103129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,2560,0.08767555819617377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,3072,0.11418399545881484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,2048,0.07236888673570421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,1536,0.026734222968419392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,2560,0.09797600242826675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,1536,0.057257778114742704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,768,0.020595555504163105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,2048,0.08163377973768446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,1024,0.022967111733224656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,1024,0.04271999994913737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,1536,0.06650311417049833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,768,0.02056533263789283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,768,0.03482310970624288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,1024,0.049775110350714795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,256,0.012098666694429187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,512,0.01886666648917728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,512,0.02725244396262699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,768,0.041703999042510986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,128,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,256,0.02337155573897892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,512,0.03365600109100342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,256,0.025082666012975905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,128,0.021665778425004747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,128,0.01753777762254079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,64,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,32,0.009007111191749573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,1024,0.027144889036814373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,128,0.01995733380317688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,64,0.02117066747612423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,32,0.020974222156736586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,512,0.01629155543115404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,256,0.017608000172509086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,65536,0.6424515512254503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,51200,1.0247981813218858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,65536,1.383318265279134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,16384,0.35178666644626194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,51200,0.5212391217549642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,16384,0.4755057758755154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,16384,0.1748622258504232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,12288,0.2864089012145996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,51200,1.53565247853597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,65536,2.077996359931098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,12288,0.13951999611324734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,65536,1.9977376725938585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,16384,0.4556933508978949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,12288,0.3303448888990614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,8192,0.1902488867441813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,10240,0.290828439924452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,10240,0.11078044441011216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,51200,1.5329511430528429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,12288,0.35120354758368594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,7168,0.1586595508787367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,8192,0.0831164452764723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,8192,0.21893155574798584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,10240,0.29752087593078613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,6144,0.13344266679551867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,7168,0.07439288828108045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,7168,0.20560710959964326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,5120,0.10982400178909302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,8192,0.2406817807091607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,6144,0.06565244330300225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,7168,0.2106320063273112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,6144,0.16671289338005912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,4096,0.09209777911504109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,5120,0.057279109954833984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,10240,0.23588177892896864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,3584,0.0755573312441508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,6144,0.18304177125295004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,4096,0.04936000042491489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,4096,0.11316711372799343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,3072,0.06814933485454984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,3584,0.04328355524275038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,5120,0.15563200579749212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,3584,0.09945866796705459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,4096,0.12882667117648655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,3072,0.0384808878103892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,3072,0.08683911297056411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,3584,0.11282844013637966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,2048,0.046316444873809814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,2560,0.03333244389957852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,2560,0.07437866926193237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,3072,0.10024444262186687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,1536,0.0343466666009691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,2048,0.02836444311671787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,2560,0.08671289020114475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,2048,0.06152977546056112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,1024,0.023989333046807185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,1536,0.02403822210099962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,1536,0.04902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,2048,0.07136178016662598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,768,0.01885422236389584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,5120,0.14044266276889378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,1024,0.03620533479584588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,1024,0.01999022232161628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,512,0.013890667094124688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,768,0.018575999471876357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,1024,0.04389155573315091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,768,0.029912001556820337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,256,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,2560,0.06053955687416924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,768,0.0359191099802653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,128,0.008593777815500895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,256,0.020021332634819877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,256,0.016175111134847004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,512,0.02905244297451443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,64,0.00794666674402025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,256,0.023299554983774822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,128,0.015839111473825242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,128,0.018583110637134977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,64,0.017895999881956313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,128,0.018535999788178336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,32,0.017895110779338412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,1536,0.05780888928307427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,512,0.01703733371363746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,65536,0.9341333177354602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,65536,0.5341164271036783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,32,0.008091555701361762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,51200,0.883204460144043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,16384,0.3127937846713596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,51200,0.4174799919128418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,65536,1.6717973285251195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,512,0.023450666003757056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,16384,0.13849243852827284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,51200,1.2117848926120334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,65536,1.7391635047064886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,12288,0.21345244513617623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,16384,0.44779290093315977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,12288,0.10630400313271417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,10240,0.18026488357120088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,12288,0.27613690164354115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,51200,1.3215182622273762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,8192,0.14292532867855495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,10240,0.08985333310233222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,10240,0.22762576738993326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,12288,0.33486755688985187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,7168,0.117740445666843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,8192,0.08120622237523396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,8192,0.18167466587490508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,6144,0.10764800177680121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,10240,0.2662746641370985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,7168,0.16059733761681452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,8192,0.21640266312493217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,6144,0.056418663925594754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,6144,0.13806222544776067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,7168,0.1928080055448744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,4096,0.06998133659362793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,5120,0.11629777484469944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,5120,0.04969066712591383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,16384,0.43106667200724286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,4096,0.04270577761862013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,5120,0.14130400286780462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,4096,0.09409422344631618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,3584,0.05965155363082886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,3072,0.05375022358364529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,3584,0.03862222366862827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,4096,0.1171591149436103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,3584,0.08228977521260579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,7168,0.06504888667000665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,3072,0.07213955455356173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,2560,0.04337244563632541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,3584,0.10312444633907741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,3072,0.03286577926741706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,2048,0.03642222285270691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,5120,0.09376000033484565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,2560,0.02860266632503933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,3072,0.09056533045238918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,1536,0.025290666355027094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,2560,0.06171022521124946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,2048,0.024500444531440735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,2048,0.0513555539978875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,1024,0.01864177816443973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,6144,0.1667635573281182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,1536,0.04137066668934292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,768,0.015107555521859063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,1024,0.018391110830836825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,1024,0.030441777573691473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,1536,0.05138755506939358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,512,0.011522666447692446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,768,0.017027555240525138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,1024,0.03849066628350152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,768,0.025059555967648823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,256,0.009401777552233802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,512,0.015722667177518208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,512,0.01973155637582143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,768,0.03277333246337043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,128,0.008543999658690559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,256,0.01479555500878228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,256,0.01683288812637329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,512,0.026575111680560645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,64,0.008261333737108443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,128,0.014833778142929077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,128,0.015491555134455362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,256,0.020249777370029025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,32,0.007688889073001013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,2560,0.0773600008752611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,64,0.01482488877243466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,128,0.016866667403115165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,2048,0.06479289134343465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,32,0.014516444669829475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,65536,0.7232106526692709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,65536,0.4250284565819635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,51200,0.5538311004638672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,16384,0.17428178257412383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,51200,0.3657191064622667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,51200,0.8770186636182996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,65536,1.2053919898139107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,16384,0.31700534290737575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,1536,0.02115733259254032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,12288,0.12842844592200384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,16384,0.1152506669362386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,12288,0.08442311154471503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,51200,1.0837937461005318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,16384,0.3581395679050022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,12288,0.2217031055026584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,65536,1.3766159481472438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,8192,0.09252711137135823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,10240,0.07746755414538913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,10240,0.194977773560418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,12288,0.2613137827979194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,7168,0.0802417794863383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,8192,0.14454400539398193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,10240,0.22244355413648817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,6144,0.0666382246547275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,8192,0.18077600002288818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,7168,0.053058667315377124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,7168,0.1269137726889716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,5120,0.05923910935719808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,6144,0.046967999802695386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,7168,0.16127288341522217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,6144,0.10917866892284817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,4096,0.0432240002685123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,10240,0.11800532870822483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,5120,0.04107111030154758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,6144,0.1411422226164076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,3584,0.03851288888189528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,4096,0.03428444597456191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,5120,0.11825954914093018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,4096,0.0747786694102817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,3072,0.03580266568395827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,3584,0.029723554849624634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,4096,0.09757866462071736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,8192,0.059119999408721924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,2560,0.02847822176085578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,3072,0.05760266383488973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,3072,0.025402666793929204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,3584,0.08661333057615493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,2048,0.024691555235120986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,2560,0.022385777698622808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,2560,0.049124442868762545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,3072,0.07526933484607272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,1536,0.018575110369258456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,2048,0.04093777802255418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,2048,0.020216888851589628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,2560,0.06464711162779066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,5120,0.09208977884716457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,1024,0.013648889131016202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,1536,0.03277955452601115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,1536,0.01791555517249637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,2048,0.054570668273501925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,768,0.011352000137170156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,1024,0.015812445018026564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,512,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,1024,0.024030221833123103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,768,0.014855111638704935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,1536,0.043543110291163124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,256,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,768,0.020244444410006206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,1024,0.032864888509114586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,512,0.013787555197874704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,128,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,512,0.016303999556435477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,768,0.027464000715149775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,256,0.013080889152155982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,64,0.00573333352804184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,256,0.013432000246312885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,32,0.005542222410440445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,512,0.02239199976126353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,128,0.013112000293201871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,128,0.012114666402339935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,256,0.017550221747822232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,64,0.011702222128709158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,128,0.01354311075475481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,32,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,65536,0.5542435646057129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,65536,0.40822309917873806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,16384,0.13715910911560059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,3584,0.06592800219853719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,51200,0.37836533122592503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,65536,0.9560818142361112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,51200,0.7562630971272787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,12288,0.11000177595350479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,16384,0.2763955593109131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,16384,0.11429333686828613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,10240,0.09438933266533746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,16384,0.3099191188812256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,12288,0.08480799860424465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,12288,0.18938933478461373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,8192,0.0708915525012546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,10240,0.0739226672384474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,51200,0.43465243445502383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,12288,0.23864711655510795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,10240,0.15743645032246908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,7168,0.06548977560467191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,8192,0.05711555480957031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,10240,0.19940000110202363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,6144,0.053139554129706494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,7168,0.11117777559492324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,7168,0.051741331815719604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,8192,0.16193332937028673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,5120,0.04777955677774218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,6144,0.04577777783075968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,7168,0.14396977424621582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,6144,0.0961831079588996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,4096,0.034597333934572004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,5120,0.03983822133806016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,51200,0.9464613596598307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,5120,0.08071911334991455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,6144,0.12404444482591416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,3584,0.030759112702475652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,4096,0.065829336643219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,5120,0.1066160003344218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,4096,0.03272711237271627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,3072,0.02843911117977566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,3584,0.02843466732237074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,3584,0.05788089169396294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,8192,0.1273288859261407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,2560,0.02335022224320306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,3072,0.05061422122849358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,3072,0.024559110403060913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,3584,0.07759288946787517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,2048,0.020005333754751418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,2560,0.02236000034544203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,2560,0.04377511143684387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,3072,0.0677679975827535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,1536,0.015541333291265698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,2048,0.03641866644223531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,2048,0.019670221540662978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,2560,0.05798400110668606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,1024,0.01144088887506061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,1536,0.017613333132531907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,1536,0.02888622217708164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,2048,0.04822133315934075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,768,0.009315555294354757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,1024,0.02164711058139801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,1536,0.03894044293297662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,4096,0.08806488911310832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,768,0.017955554856194388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,768,0.014819555812411837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,512,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,512,0.013734222286277346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,256,0.00564533347884814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,512,0.01442399952146742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,512,0.021593777669800654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,256,0.011799111134476133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,128,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,256,0.013110222087966071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,65536,1.2142933739556205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,64,0.006436444405052397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,256,0.016499555773205228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,128,0.013089777694808112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,32,0.004567999806669024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,128,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,1024,0.01574222246805827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,64,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,32,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,1024,0.029267556137508813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,51200,0.3048871093326145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,768,0.025053333905008104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,65536,0.31933776537577313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,16384,0.09762577878104316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,65536,0.8003582424587674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,51200,0.2540880044301351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,51200,0.6320364210340712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,16384,0.21558933787875703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,16384,0.08393510844972397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,12288,0.07209510935677423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,128,0.012949333422713809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,65536,1.05949158138699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,10240,0.062239110469818115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,12288,0.06732444630728827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,65536,0.405312008327908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,16384,0.2745235496097141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,12288,0.16153155432807073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,8192,0.04889333248138428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,10240,0.0596675541665819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,7168,0.041900446017583214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,12288,0.20888177553812662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,10240,0.131312886873881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,8192,0.10522933138741387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,6144,0.038591110044055514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,7168,0.041371554136276245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,7168,0.09318311346901788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,6144,0.03671288821432326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,7168,0.12762399514516196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,6144,0.08041066593594022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,5120,0.030431999100579157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,6144,0.11159022649129231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,5120,0.03229066729545593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,4096,0.023927999867333308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,5120,0.06808355781767103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,5120,0.09479200177722508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,4096,0.05537066525883145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,3584,0.021187555458810594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,8192,0.04559200008710226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,4096,0.07812888754738702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,3072,0.018703111343913607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,10240,0.17620266808403862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,3584,0.04862222075462341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,8192,0.1438479953342014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,2560,0.01662488944000668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,3072,0.019291554888089497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,3584,0.06828088892830743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,3072,0.04310577776696947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,2048,0.013585777746306526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,2560,0.017511111166742113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,3072,0.059756444560156934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,2560,0.03677422139379714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,2048,0.03047555685043335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,1536,0.011057777537239922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,4096,0.024940444363488093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,2048,0.042465776205062866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,1536,0.02440711193614536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,1024,0.008291555775536431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,1536,0.014467555615637036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,3584,0.021272889441914026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,1536,0.033987555238935686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,1024,0.01314399970902337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,768,0.0070088886552386815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,1024,0.025984888275464375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,768,0.015511110424995422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,512,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,768,0.02233244478702545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,2560,0.051169776254230075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,256,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,512,0.012444444828563266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,512,0.012019555601808759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,51200,0.8357004589504666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,128,0.0038968887594011095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,256,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,256,0.010030222435792288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,512,0.0189057770702574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,64,0.003519999898142285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,128,0.008976000050703684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,128,0.011336888704035016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,256,0.014479110638300577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,1024,0.01830844415558709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,64,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,32,0.008040889269775814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,65536,0.2429626782735189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,768,0.01274400038851632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,51200,0.19455111026763916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,2048,0.0159244438012441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,65536,0.31240177154541016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,16384,0.05870133638381958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,65536,0.6623813311258951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,51200,0.2486808829837375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,51200,0.5186088879903158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,32,0.0038888889054457345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,16384,0.08329600095748901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,12288,0.04645688997374641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,16384,0.172069337632921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,10240,0.036919110351138644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,128,0.011677333050303988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,65536,0.9030177858140734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,12288,0.06606666909323798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,16384,0.23604000939263237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,8192,0.031383110417260066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,12288,0.1311697827445136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,51200,0.7133039898342557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,10240,0.05862666500939263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,7168,0.02700000007947286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,10240,0.11031822363535564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,8192,0.04474222328927782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,12288,0.18038222524854872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,6144,0.02309155629740821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,8192,0.08979199992285834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,10240,0.15299555990431044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,7168,0.04050488935576545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,7168,0.07889866828918457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,5120,0.016398222910033334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,6144,0.036235554350747004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,8192,0.12519555621676975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,4096,0.013566222455766467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,6144,0.06804711288876004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,7168,0.10967199669943915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,3584,0.012331555287043253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,5120,0.05708177884419759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,4096,0.023749333288934495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,6144,0.09720800320307414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,3072,0.01109955542617374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,5120,0.08243111107084486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,3584,0.021029333273569744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,3584,0.04115555683771769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,2560,0.009686222506894005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,4096,0.06718844175338745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,3072,0.019154666198624503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,3072,0.036166220903396606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,2048,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,3584,0.05893066856596205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,2560,0.017599110802014668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,2560,0.03065955638885498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,1536,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,3072,0.05222488774193657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,2048,0.01608711149957445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,2560,0.04473955432573954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,1024,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,1536,0.014717333846622042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,2048,0.03683111071586609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,1536,0.02074044446150462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,768,0.004642666627963384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,5120,0.031758222315046526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,1024,0.013183111117945777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,1536,0.029653333955340918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,512,0.00416355538699362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,4096,0.046295109722349376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,768,0.013152889079517789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,768,0.012726222475369772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,1024,0.02292977770169576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,256,0.003571555432346132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,512,0.01202933324707879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,512,0.01075644459989336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,768,0.01960533360640208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,128,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,256,0.011384888655609555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,256,0.008560889297061497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,512,0.016526222229003906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,64,0.002826666666401757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,128,0.0070293330483966405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,128,0.01144444445768992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,32,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,256,0.012763555679056378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,2048,0.025754666990704004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,64,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,128,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,32,0.006648889018429651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,65536,0.1775511105855306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,1024,0.01575911045074463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,65536,0.32098044289482963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,16384,0.04547377758555942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,65536,0.5907466676500108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,51200,0.25195378727383083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,51200,0.4611457718743219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,12288,0.035792890522215105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,16384,0.08461155494054158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,16384,0.1541164451175266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,65536,0.8844533496432834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,10240,0.029475556479560003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,12288,0.11750488811069065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,12288,0.06850222084257337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,16384,0.23075289196438262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,51200,0.6931226518419055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,8192,0.025666667355431452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,10240,0.06090933084487915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,51200,0.14067110750410292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,12288,0.17706667052374947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,7168,0.021339555581410725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,10240,0.09877777761883205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,6144,0.017268444101015728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,8192,0.04526488979657491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,8192,0.0796639985508389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,7168,0.041519999504089355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,5120,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,7168,0.07006755802366468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,6144,0.03736266824934218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,8192,0.12292622195349799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,6144,0.06044444772932264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,4096,0.009136888715955947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,7168,0.10920888847774929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,5120,0.031982223192850746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,5120,0.050741334756215416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,3584,0.008301333420806462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,6144,0.09418044487635295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,4096,0.024162666665183172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,4096,0.041273779339260526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,3072,0.007605333295133378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,3584,0.02165511084927453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,5120,0.08100355333752103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,3584,0.036333332459131874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,4096,0.06519110997517903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,2560,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,3072,0.01960355540116628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,3072,0.031930665175120033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,2048,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,2560,0.0181057784292433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,2560,0.027228444814682007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,3072,0.05031288994683159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,1536,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,2048,0.016680889659457736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,2048,0.022852443986468848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,2560,0.04347199863857693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,1024,0.004233777936961916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,10240,0.15018310811784533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,1536,0.014826666977670459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,1536,0.018423999349276226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,2048,0.03537599907981025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,1024,0.013464888764752282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,1024,0.014112888111008538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,1536,0.028943998946083918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,512,0.0031946665710873077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,768,0.012783110969596438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,768,0.012053333222866058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,1024,0.02201066745652093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,256,0.002826666666401757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,512,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,768,0.01921955578856998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,512,0.009718221922715506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,128,0.002481777841846148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,256,0.007810666329330868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,256,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,512,0.016132444143295288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,3584,0.057486222849951855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,128,0.011360000405046674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,32,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,128,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,64,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,32,0.006032889087994893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,65536,0.16520088248782688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,768,0.003588444242874781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,16384,0.04790577623579237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,65536,0.561538643307156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,12288,0.03459022111362881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,16384,0.14616533120473227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,51200,0.4409004317389594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,10240,0.028384000062942505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,8192,0.023983110984166462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,64,0.0024746666765875286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,7168,0.02089511189195845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,12288,0.11026844713422988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,256,0.01242844429281023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,6144,0.015295111470752291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,5120,0.00982488857375251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,10240,0.09193778038024902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,4096,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,7168,0.06544444296095106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,6144,0.05665333403481377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,3584,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,3072,0.006659555352396435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,5120,0.04787733488612705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,2560,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,4096,0.039127998881869845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,2048,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,3584,0.03455555438995361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,1536,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,3072,0.0302782224284278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,1024,0.0041555555330382455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,2560,0.025986666480700176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,768,0.003549333247873518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,2048,0.021864889396561518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,512,0.0032142222755485107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,1536,0.0177875558535258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,256,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,1024,0.01365066650840971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,128,0.0022026666750510535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,768,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,64,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,512,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,32,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,256,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,8192,0.07453866799672444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,128,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,64,0.005789333333571752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,51200,0.13185332881079778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,32,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,65536,0.17292977703942192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,12288,0.03457866774664985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,16384,0.04491110973887973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,10240,0.026885333988401625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,8192,0.022637334134843614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,16384,0.14646844069163004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,12288,0.10999200079176162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,7168,0.019719999697473314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,51200,0.4403226640489366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,6144,0.014929778046078153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,10240,0.09183644586139256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,65536,0.5615999963548448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,8192,0.07426133420732287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,4096,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,3584,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,7168,0.06545866860283746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,3072,0.006001777946949005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,6144,0.056695110268063016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,5120,0.04770933257208931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,2560,0.005324444423119227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,2048,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,3584,0.036766221125920616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,1536,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,3072,0.032122668292787336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,2560,0.027123555541038513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,1024,0.0038480001191298165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,768,0.0035128887328836652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,2048,0.02264533274703556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,512,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,51200,0.12535022364722356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,256,0.00254844439526399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,1024,0.013932444983058505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,768,0.011752000285519494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,128,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,64,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,512,0.00943199959066179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,32,0.0021537778278191886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,256,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,5120,0.00905777762333552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,128,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,64,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,32,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,4096,0.0391813317934672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,1536,0.018441778090265062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,16384,1.9122435251871746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,12288,4.650211758083768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,16384,5.9817352294921875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,16384,6.39211188422309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,12288,1.5126683976915147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,12288,4.731958177354601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,51200,7.434607611762153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,10240,1.1750675837198894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,10240,3.9167039659288196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,8192,0.9627493752373589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,10240,3.6348410712348094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,8192,2.944745805528429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,7168,0.8232800165812174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,8192,2.8965945773654513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,7168,2.4056479136149087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,7168,2.9570240444607205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,6144,0.7032088703579373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,6144,2.0849759843614364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,6144,2.1343165503607855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,5120,0.6019475724962022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,5120,1.8620391421847875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,4096,0.5128222041659886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,5120,1.7265733083089192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,4096,1.5919680065578883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,4096,1.4237502415974934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,3584,0.43081243832906085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,3584,1.1991857952541776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,3584,1.2575129403008354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,3072,0.3845591015285916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,3072,1.154762691921658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,128,0.009670221971140968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,3072,1.0567004945543077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,2560,0.870717313554552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,2560,0.3476044336954753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,2048,0.28066221872965497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,2560,0.8966400358412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,1536,0.5593928760952419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,2048,0.7046453158060709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,2048,0.704528914557563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,1536,0.2359093295203315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,1024,0.20524088541666666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,1536,0.5857502089606391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,1024,0.402129782570733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,768,0.28284621238708496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,512,0.2312871085272895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,768,0.1647333304087321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,1024,0.47086132897271055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,768,0.34398576948377824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,512,0.14242311318715414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,256,0.17539111773173013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,51200,18.76383293999566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,256,0.12485867076449925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,256,0.2590346601274279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,128,0.24541778034634062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,32,0.10848177803887261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,64,0.10524266958236694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,64,0.2468195491366916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,32,0.25444356600443524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,128,0.1204035547044542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,128,0.15242577923668757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,512,0.3692639933692084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,16384,4.345950232611762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,51200,5.729736751980251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,51200,27.341620551215275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,65536,7.565330505371094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,16384,1.4449031617906358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,12288,1.0055884255303276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,12288,3.340125401814779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,16384,5.006476508246528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,12288,3.5018249087863498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,10240,0.9112311469184028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,51200,13.863909403483072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,10240,3.2152498033311634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,10240,2.7940461900499134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,8192,2.2994942135281033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,8192,0.7034657796223959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,8192,2.242953830295139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,7168,0.6282400025261773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,7168,2.186356438530816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,7168,1.8688312106662328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,6144,0.5545768737792969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,6144,1.7035591337415907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,6144,1.9580150180392797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,5120,0.4721973207261827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,65536,17.699595133463543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,5120,1.3452391094631617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,4096,0.38027199109395343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,5120,1.3216817643907335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,4096,1.0976044336954753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,3584,0.9615635342068143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,3584,0.372524446911282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,4096,1.1179564793904622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,3072,0.8066870901319715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,3584,0.9431289037068685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,2560,0.7307519912719727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,3072,0.30040889316134983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,3072,0.8201368649800619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,2560,0.2705964512295193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,2560,0.6882150967915853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,2048,0.22101599640316436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,1536,0.38743821779886883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,1536,0.18148267269134521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,1024,0.31059466467963326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,1536,0.45607911215888125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,1024,0.14216355482737222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,2048,0.6470915476481119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,768,0.2506871223449707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,1024,0.3474133279588487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,512,0.17170400089687773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,768,0.1246417760848999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,256,0.1188080045912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,768,0.2604657808939616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,512,0.11035111215379502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,2048,0.533531559838189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,128,0.09310844209459092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,512,0.29255199432373047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,256,0.09552088710996841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,64,0.09050577878952026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,256,0.20022043916914198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,128,0.09639110830095078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,32,0.09573600027296279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,128,0.1898346609539456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,64,0.19476355446709526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,32,0.19781777593824598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,51200,21.80615743001302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,65536,2.3388239542643228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,16384,1.3988978068033855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,51200,1.9742355346679688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,16384,0.4569270875718858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,51200,4.32932874891493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,12288,1.3059671190049913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,16384,1.7261724472045898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,12288,0.35040977266099715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,12288,1.1978372997707791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,65536,5.7002614339192705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,10240,0.2931111123826769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,10240,1.0112693574693468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,8192,0.6835155487060547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,10240,0.9505173365275065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,7168,0.21119377348158094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,7168,0.7893830935160319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,8192,0.7331813176472982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,7168,0.650262196858724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,6144,0.18901333543989393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,6144,0.5895599789089626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,8192,0.2358835538228353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,5120,0.4126106633080377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,6144,0.6419902377658421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,5120,0.205112894376119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,5120,0.4433111084832086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,4096,0.12968622313605413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,4096,0.35001065995958114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,3584,0.3267751004960802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,3584,0.11651289463043213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,4096,0.3719164530436198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,3072,0.27207199732462567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,3584,0.3150613307952881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,2560,0.19841955767737496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,3072,0.3175475597381592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,2560,0.0935351120101081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,51200,8.533114115397135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,2560,0.29924265543619794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,1536,0.1474799977408515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,2048,0.19203644328647188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,1536,0.06832089026769002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,1536,0.15492799546983507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,1024,0.09309333562850952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,768,0.0776853362719218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,1024,0.05184444453981188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,1024,0.1162373357348972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,768,0.04695377747217814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,512,0.05968799855973986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,3072,0.10309866401884292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,256,0.039120889372295804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,512,0.043876445955700345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,512,0.07654044363233778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,128,0.028705779049131606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,256,0.038375112745496966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,256,0.06881066825654772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,2048,0.07890666855706109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,64,0.025469332933425903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,32,0.026019554999139573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,128,0.0380311111609141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,128,0.0649777783287896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,65536,10.661615159776476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,64,0.06646755668852064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,32,0.06749866406122844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,768,0.09175022443135579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,2048,0.16377955012851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,65536,33.41026136610243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,16384,1.008148405287001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,51200,1.2629342608981662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,65536,1.8983342912462022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,16384,0.35056267844306094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,16384,1.277905782063802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,12288,0.9694835874769422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,51200,3.3010826110839844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,10240,0.7881849077012805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,12288,1.103851530287001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,10240,0.22910934024386934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,12288,0.2797093391418457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,8192,0.18404711617363823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,10240,0.7246426476372613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,8192,0.6373644404941136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,65536,4.517604404025607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,8192,0.5610399776034886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,7168,0.16287822193569607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,7168,0.444543997446696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,6144,0.14269333415561253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,6144,0.4237111144595676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,7168,0.4878337648179796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,5120,0.329839997821384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,5120,0.13337688975863987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,4096,0.2609653207990858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,5120,0.36629509925842285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,4096,0.10087111261155869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,3584,0.2285146713256836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,4096,0.305731561448839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,3584,0.08967200252744888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,3072,0.21015200350019667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,3584,0.24111111958821616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,3072,0.07987822426689996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,2560,0.17596711052788627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,51200,6.064542982313369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,3072,0.21205243799421522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,2048,0.1447519991132948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,2560,0.07047733333375719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,2560,0.17555556032392713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,2048,0.06016800138685438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,1536,0.10821511348088582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,1024,0.07969599962234497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,1536,0.04938311047024197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,6144,0.44362934430440265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,2048,0.16899822817908394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,768,0.061653335889180504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,1536,0.11472354994879828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,1024,0.04145066605673896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,512,0.0461715559164683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,256,0.028731554746627808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,1024,0.11189422342512344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,768,0.04138844543033176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,768,0.06947466399934557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,512,0.034916443957222834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,128,0.026416889495319788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,64,0.02093155516518487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,256,0.03821333249409994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,65536,7.293876647949219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,256,0.053563555081685386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,128,0.031467556953430176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,512,0.05927822324964735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,128,0.05061422122849358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,64,0.05164711011780632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,32,0.055904891755845815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,32,0.021176889538764954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,16384,0.8638568984137641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,51200,1.0861226187811959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,65536,1.4702195061577692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,16384,0.29621778594122994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,12288,0.22352888849046496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,16384,0.9896177715725369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,12288,0.9616070853339301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,12288,0.714537779490153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,51200,2.911483552720812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,10240,0.19617511166466606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,10240,0.5983840094672309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,8192,0.15744711293114558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,65536,3.4753661685519748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,8192,0.5077804459465874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,7168,0.4048782189687093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,7168,0.13826310634613037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,8192,0.5268595483567979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,6144,0.39496533075968426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,7168,0.411626656850179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,6144,0.12149155139923096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,5120,0.2789075639512804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,5120,0.10429955853356256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,10240,0.6337182256910536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,4096,0.2181831068462796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,6144,0.40060265858968097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,51200,4.239217970106337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,3584,0.2037093374464247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,5120,0.4321271048651801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,3072,0.1639839940600925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,4096,0.26316711637708873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,3584,0.0780151089032491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,2560,0.13398310873243544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,3072,0.0773093303044637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,2560,0.06131911277770996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,2048,0.11709866258833145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,3072,0.1746764447953966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,2560,0.1571848922305637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,1536,0.09392622444364761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,2048,0.052404443422953285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,1024,0.06245333618587918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,2048,0.12318489286634658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,1536,0.04452000061670939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,768,0.04992799957593282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,1536,0.09785599841011895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,1024,0.03951644566324022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,512,0.0339680016040802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,1024,0.07482310798433092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,768,0.033624887466430664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,256,0.02466222147146861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,512,0.03073244293530782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,768,0.05983733468585544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,128,0.021497777766651575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,512,0.053188443183898926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,256,0.027797334724002417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,4096,0.08687288893593682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,256,0.045441776514053345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,128,0.027262222435739305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,65536,6.236500634087457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,128,0.04305600126584371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,3584,0.23002844386630586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,64,0.04395555456479391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,32,0.04553066690762838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,64,0.019144889381196763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,32,0.017993777990341187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,65536,1.3792044321695964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,16384,0.7572239769829644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,51200,2.2310337490505643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,16384,0.2918391227722168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,12288,0.4712906413608127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,65536,2.860061433580187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,12288,0.20207111040751138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,16384,1.0761138068305123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,10240,0.43106312221950954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,12288,0.5998346540662978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,10240,0.16626755396525064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,8192,0.1339226696226332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,8192,0.38873688379923504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,7168,0.3305235438876682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,51200,1.0122017330593533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,10240,0.4886915418836806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,51200,3.510700437757704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,7168,0.11673688888549805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,6144,0.10281333658430312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,6144,0.2853306664360894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,7168,0.3262080086602105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,6144,0.281713777118259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,5120,0.08887022071414524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,4096,0.22501778602600098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,5120,0.2444817754957411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,3584,0.15943200058407253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,4096,0.07410666677686903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,3072,0.13933689064449734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,3584,0.06622133652369182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,3584,0.2431253327263726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,2560,0.11192977428436279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,3072,0.05894577503204346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,3072,0.1431528859668308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,2048,0.095287111070421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,2560,0.05172533459133572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,1536,0.07192711035410564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,8192,0.3974657853444417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,65536,5.416395399305556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,5120,0.21658133135901558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,1024,0.05263466636339823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,1536,0.036441776487562395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,1536,0.08098933431837294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,768,0.03756266501214769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,1024,0.034846223062939115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,512,0.029334220621320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,1024,0.062041779359181724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,768,0.028404444456100464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,256,0.0196426659822464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,768,0.04976355698373583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,512,0.024886222349272832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,4096,0.18601244025760225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,128,0.015424888994958667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,512,0.04164444406827291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,64,0.013389333254761167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,256,0.023007111416922674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,256,0.036473777559068464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,32,0.013723555538389417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,128,0.03410755594571432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,64,0.03439911206563314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,32,0.03490133417977227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,2560,0.17319999800788033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,2048,0.04830933279461331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,2048,0.10141866736941868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,16384,0.5381662050882975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,128,0.02267644471592373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,51200,0.8416159947713217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,65536,1.1768791410658095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,16384,0.2353937890794542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,12288,0.17214132679833305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,16384,0.7424986627366809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,12288,0.5364658037821451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,51200,1.838181389702691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,65536,2.35117064581977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,10240,0.41011643409729004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,12288,0.5105128818088108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,10240,0.18155554930369058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,8192,0.3309182325998942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,10240,0.4240729014078776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,8192,0.13779466682010227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,7168,0.3265937699211968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,7168,0.11562844117482503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,6144,0.2239209016164144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,8192,0.3539902104271783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,7168,0.3222595585717095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,6144,0.0953777763578627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,5120,0.19374667273627388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,4096,0.16574045022328696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,5120,0.07963733540640937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,6144,0.26433777809143066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,4096,0.07052977879842122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,5120,0.2352986600663927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,3584,0.05954222546683418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,3072,0.1184542179107666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,4096,0.21761155128479004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,3584,0.146978669696384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,51200,3.4261182149251304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,2560,0.0961644450823466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,3072,0.05310044354862637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,2048,0.08445600006315444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,3072,0.12960710790422228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,2560,0.0535697771443261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,1536,0.0636062224706014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,2560,0.11025955279668172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,1024,0.0432515541712443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,2048,0.09193155500623916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,1536,0.03312355611059401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,65536,4.254820505777995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,768,0.03219377663400438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,1536,0.07744533485836454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,512,0.024918221765094336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,1024,0.027424888478385076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,768,0.024842666255103216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,1024,0.05487111210823059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,256,0.017664000391960144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,768,0.04413955410321554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,128,0.014757333530320061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,3584,0.13452000088161892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,512,0.036109334892696805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,256,0.020495999190542433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,64,0.012381333443853589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,256,0.031073778867721558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,32,0.012480888929631976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,128,0.030064887470669214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,64,0.029993779129452173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,32,0.03000977635383606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,2048,0.04222666554980808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,512,0.022368000613318548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,128,0.020287111401557922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,65536,0.947697745429145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,16384,0.5436426798502604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,51200,1.6123972998725042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,51200,0.7316524187723795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,65536,2.169067594740126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,16384,0.20406222343444824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,12288,0.15133333206176758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,12288,0.4042106734381781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,10240,0.33304177390204537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,16384,0.946091545952691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,10240,0.12754578060574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,8192,0.28259645567999947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,12288,0.5837857988145616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,8192,0.10223733054267035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,10240,0.38722843594021267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,7168,0.2501902315351698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,7168,0.09210311041937934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,8192,0.3208808898925781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,6144,0.08034133248858981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,51200,2.8619058397081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,7168,0.2857031027475993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,6144,0.22228709856669107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,5120,0.07250133487913343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,4096,0.1373786661359999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,65536,3.422155592176649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,5120,0.18707911173502603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,3584,0.11600266562567817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,4096,0.058911111619737416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,3072,0.09839822186364068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,4096,0.148960895008511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,3584,0.05791999896367391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,2560,0.08936711152394612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,3072,0.04641511042912801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,3584,0.15278311570485434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,2048,0.06945421960618761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,2560,0.04339111182424757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,3072,0.11332889397939046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,1536,0.053936888774236046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,1024,0.03966310951444838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,2560,0.09824711084365845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,1536,0.0291920006275177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,2048,0.0795324444770813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,768,0.028239998552534316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,1024,0.025335111551814612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,1536,0.06473155816396077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,6144,0.21666577127244738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,1024,0.048208889034059316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,768,0.022117333279715642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,256,0.015494222442309061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,512,0.019952888290087383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,5120,0.16831110583411324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,768,0.03913955556021796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,128,0.014261333478821648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,512,0.031622221072514854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,256,0.018236445056067575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,64,0.01464000013139513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,256,0.02718577782313029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,32,0.015450666348139444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,128,0.02641955514748891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,64,0.026100445124838088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,32,0.026014222039116755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,2048,0.036064889695909284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,512,0.0211208893193139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,65536,0.7792764239841037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,128,0.018200000127156574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,51200,0.5986026657952203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,51200,1.3272292878892686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,65536,1.7039946450127497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,16384,0.19480621814727783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,16384,0.5910879770914713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,12288,0.339224894841512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,16384,0.527508470747206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,12288,0.12899111376868352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,10240,0.37554489241705996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,10240,0.11158666345808242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,12288,0.4079333411322699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,8192,0.2515893247392443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,8192,0.0893404417567783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,10240,0.3416897720760769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,51200,2.269625769721137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,8192,0.33748798900180393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,6144,0.1819324493408203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,7168,0.08011466926998563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,7168,0.2865617805057102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,5120,0.15045155419243708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,65536,3.0309102800157337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,6144,0.07470311058892144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,6144,0.20067732863956025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,5120,0.06564444303512573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,4096,0.11994844012790257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,3584,0.10652177863650852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,5120,0.15739822387695312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,4096,0.05262310968505012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,3584,0.05214577913284302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,3072,0.08987644645902847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,4096,0.1342382166120741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,3072,0.039926220973332725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,2048,0.05719021956125895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,3584,0.11019644472334121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,3072,0.0970417790942722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,1536,0.043679111533694796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,2560,0.03606577714284261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,2048,0.0321084459622701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,1024,0.029358221424950495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,2560,0.08588088883294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,1536,0.02510844502184126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,768,0.02401599950260586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,7168,0.21447377734714082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,512,0.017157332764731515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,1536,0.054936889145109385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,768,0.020080000162124634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,1024,0.04161511196030511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,256,0.013316444224781461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,512,0.017614222235149808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,768,0.03423733181423611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,128,0.010934222075674268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,256,0.01644622286160787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,64,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,256,0.023573334018389385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,128,0.016400888562202454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,32,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,128,0.02269688910908169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,2560,0.0697751111454434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,64,0.022650667362742957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,32,0.022545778089099463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,2048,0.06877599822150336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,1024,0.02180888917711046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,65536,0.6037013265821669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,512,0.027225777506828308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,16384,0.33096710840861004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,51200,1.0386169221666124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,51200,0.4839475419786241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,16384,0.1402240064409044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,65536,1.4202177259657118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,12288,0.2305075592464871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,12288,0.13316444555918375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,10240,0.22943022516038683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,10240,0.0925546685854594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,16384,0.4872328970167372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,8192,0.21280977461073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,12288,0.31143911679585773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,10240,0.27418221367730033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,7168,0.17088800006442598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,8192,0.07402133279376559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,6144,0.13638667265574136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,7168,0.0675777792930603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,6144,0.058508442507849805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,65536,2.271731482611762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,7168,0.19614577293395996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,4096,0.0943137804667155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,6144,0.16058844990200466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,5120,0.05221511258019341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,5120,0.12936000029246011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,3584,0.07832888762156169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,4096,0.04870577653249105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,4096,0.10805866453382705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,3584,0.0369031098153856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,3584,0.09471733040279812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,3072,0.03386488887998793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,3072,0.08440799845589532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,2048,0.04785511228773329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,2560,0.06898755497402616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,51200,1.7571626239352758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,2048,0.025714667307005987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,2048,0.05849155452516344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,5120,0.11364889144897461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,1536,0.03357155455483331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,1024,0.022044445077578228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,1536,0.022234666678640578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,1536,0.04599644409285652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,1024,0.018632888793945312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,3072,0.07149866554472181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,1024,0.03461866577466329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,512,0.013776889277829064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,768,0.017024000485738117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,2560,0.05766666597790188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,256,0.010354667074150508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,768,0.02888266576661004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,512,0.022615111536449854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,512,0.015485333071814643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,2560,0.029721776644388836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,128,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,256,0.014504000544548035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,64,0.009342222577995723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,32,0.009665778113736046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,256,0.01973511113060845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,128,0.014735110931926303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,8192,0.24438312318589953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,64,0.018754666050275166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,32,0.018903111418088276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,768,0.018560000591807894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,65536,0.5503706402248806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,16384,0.32534488042195636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,65536,1.1395591100056965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,51200,0.44750308990478516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,51200,0.9876622094048394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,16384,0.13132888740963405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,12288,0.22200711568196616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,128,0.01925511161486308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,12288,0.13104177845848933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,10240,0.20204799705081514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,16384,0.41197866863674587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,10240,0.08687822024027507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,8192,0.18175466855367026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,8192,0.06943200031916301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,7168,0.13725422488318548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,51200,1.454927126566569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,10240,0.2836097876230876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,6144,0.10352000263002183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,8192,0.2135359976026747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,7168,0.06106399827533298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,65536,2.0819360944959855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,5120,0.09995377726025051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,7168,0.18912088871002197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,6144,0.05812266800138685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,5120,0.04657422171698677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,4096,0.08158489068349202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,6144,0.14508888456556532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,3584,0.0694720016585456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,4096,0.03826399975352817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,5120,0.12153955300649007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,3072,0.05496444304784139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,2560,0.04973777797487047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,4096,0.09777244594362046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,3584,0.08493244647979736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,2048,0.038351111941867404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,3072,0.07415022452672322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,2560,0.02788533435927497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,1536,0.0328693323665195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,2560,0.0652408864763048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,2048,0.023307555251651343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,1024,0.022812444302770827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,12288,0.28324445088704425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,2048,0.05227111114395989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,1536,0.019555555449591745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,768,0.019239111079110038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,1024,0.017208000024159748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,1536,0.04174933168623182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,512,0.015024888846609326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,1024,0.03143644332885742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,768,0.015807110402319167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,512,0.01460533340771993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,768,0.02571288910177019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,128,0.00831288927131229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,512,0.02081155611409081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,256,0.01369866645998425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,64,0.007270221908887227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,32,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,128,0.013791110780504016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,128,0.01751377847459581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,64,0.01718222267097897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,32,0.017215111189418368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,3072,0.030646221505271062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,256,0.01111288865407308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,65536,0.47809775670369464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,51200,0.7333040237426758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,65536,1.232669300503201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,16384,0.3481626775529649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,51200,0.37384266323513454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,3584,0.0398942232131958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,16384,0.12231111526489258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,12288,0.2560257646772596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,16384,0.3323591020372179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,51200,1.2046533160739474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,10240,0.18412444326612684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,12288,0.08882577551735772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,10240,0.07691555553012423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,8192,0.14960622787475586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,12288,0.29358310169643825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,7168,0.13112000624338785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,10240,0.25085777706570095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,8192,0.062459554937150746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,8192,0.16275200578901503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,6144,0.10335822237862481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,7168,0.06442222330305311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,5120,0.08435733450783624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,6144,0.047129776742723256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,256,0.018062222335073683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,4096,0.06760533650716145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,6144,0.12552711698744032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,5120,0.04025777843263414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,3584,0.060659554269578725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,4096,0.03276177909639146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,3072,0.05079022381040785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,4096,0.08404977454079522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,3584,0.0750746660762363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,65536,1.5451190736558702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,3072,0.026867555247412786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,2560,0.04653244548373752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,2048,0.03597777750757005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,3072,0.06515644656287299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,2560,0.02433066732353634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,1536,0.026519111461109583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,2560,0.05572444200515747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,2048,0.021106666988796655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,1024,0.018965333700180054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,2048,0.04692177640067207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,7168,0.1497404442893134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,1536,0.018592000007629395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,1024,0.01590399940808614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,1024,0.028132445282406274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,512,0.011418666276666852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,768,0.01499022212293413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,256,0.00851022203763326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,768,0.023431999815834895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,5120,0.11331289344363743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,512,0.013719999955760108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,128,0.007703999678293864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,3584,0.030284444491068523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,512,0.018808888064490426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,256,0.012842666771676807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,32,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,128,0.012881778180599213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,128,0.015611555841233997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,64,0.015154666370815702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,32,0.015464888678656684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,768,0.015503111812803479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,65536,0.43266667260064023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,65536,0.9356115129258898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,51200,0.636174201965332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,64,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,256,0.0162764439980189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,51200,0.3440577718946669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,16384,0.28299111790127224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,16384,0.10418133603201972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,65536,1.2592924965752494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,12288,0.1908728943930732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,16384,0.29515287611219615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,12288,0.10058044724994236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,10240,0.17039555973476836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,8192,0.12211110856797959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,10240,0.07039378086725871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,12288,0.22416978412204316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,51200,0.9469102223714193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,8192,0.05475199884838528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,7168,0.10184355576833089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,10240,0.18639111518859863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,1536,0.0376835564772288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,6144,0.0948222213321262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,7168,0.04953866534762912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,8192,0.20041244559817842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,5120,0.08093066347969903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,6144,0.04230666822857327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,4096,0.06197422080569797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,7168,0.12424266338348389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,5120,0.0358586675590939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,3584,0.05395288930998909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,6144,0.10790933503044976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,3072,0.04758844441837735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,5120,0.09102400143941243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,3584,0.026318222284317017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,2560,0.04203022188610501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,3072,0.02410488824049632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,2048,0.03374044431580438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,3072,0.057329780525631376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,2560,0.02160888910293579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,1536,0.026551110876931086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,2048,0.019189332922299702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,2560,0.048872000641293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,1024,0.019293333093325298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,2048,0.04144622219933404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,1536,0.016953777935769822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,768,0.013456000222100152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,1024,0.014819555812411837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,1024,0.02481422159406874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,512,0.010568000376224518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,768,0.013836444252067141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,768,0.020711110697852243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,256,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,512,0.012763555679056378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,512,0.01706133286158244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,256,0.012411555482281579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,128,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,4096,0.030031111505296495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,256,0.014447111222479077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,64,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,4096,0.07876177628835042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,32,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,128,0.012072000238630505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,3584,0.06548000044292875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,128,0.014135999812020196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,64,0.01346666696998808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,32,0.01347733371787601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,1536,0.03293955657217238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,65536,0.3425573243035211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,65536,0.6002524693806967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,51200,0.47591908772786456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,51200,0.2708328829871284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,12288,0.14404711458418104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,16384,0.08649155828687881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,16384,0.23768178621927896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,10240,0.11960444185468887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,12288,0.06700088580449422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,12288,0.17947467168172201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,65536,1.0047377480400932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,8192,0.10467288891474406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,10240,0.0593128866619534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,7168,0.09279111358854507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,8192,0.045817779170142285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,10240,0.15038755204942492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,6144,0.07560622029834323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,7168,0.04152889053026835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,8192,0.12113600307040745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,5120,0.06342399782604642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,7168,0.10776178042093913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,6144,0.035359111097123884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,4096,0.04489511251449585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,6144,0.09176355600357056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,5120,0.030408889055252075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,3584,0.03933155536651611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,5120,0.07793866925769381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,3072,0.03288799855444167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,4096,0.025768889321221247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,16384,0.18528800540500215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,2560,0.02937600016593933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,3584,0.023285332653257582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,4096,0.0630355543560452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,2048,0.024200000696712073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,3072,0.02130399975511763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,3584,0.0551511115498013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,2560,0.01943288909064399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,3072,0.04845066534148323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,1024,0.013773333695199756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,2048,0.017204445269372728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,51200,0.7940186924404569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,1536,0.0155404441886478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,768,0.011081777513027191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,1024,0.013394667042626275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,512,0.008575110799736446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,1024,0.02176888949341244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,768,0.018237334158685472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,256,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,512,0.011741333537631564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,512,0.014929778046078153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,128,0.006127111199829314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,256,0.011054221954610614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,256,0.012757333616415659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,64,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,128,0.011115555961926779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,32,0.006063999815119638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,128,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,1536,0.01860000027550591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,64,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,2560,0.04183822207980686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,32,0.011734222372372946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,2048,0.03546844588385688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,1536,0.028668443361918133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,768,0.012764444781674279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,65536,0.3039582305484348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,16384,0.1406737830903795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,51200,0.41556623246934676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,51200,0.24640711148579916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,12288,0.1086462206310696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,16384,0.07686844136979845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,10240,0.09999644756317139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,12288,0.06341599755816989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,16384,0.19591732819875082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,65536,0.8050479888916016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,51200,0.6268329090542263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,8192,0.07686577902899848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,12288,0.15904533863067627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,10240,0.05532977647251553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,7168,0.06369600031110975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,10240,0.12528799639807806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,6144,0.052998221582836576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,8192,0.04305333230230543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,7168,0.03894666830698649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,8192,0.10133333338631524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,6144,0.03331999977429708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,4096,0.035988443427615695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,5120,0.028219554159376357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,5120,0.06589155726962619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,4096,0.023847111397319373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,3584,0.03264177838961283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,65536,0.5357839796278212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,4096,0.053492443429099194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,3584,0.021990221407678392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,3584,0.0477128889825609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,2560,0.022711111439598933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,3072,0.04197066691186693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,2560,0.018241778016090393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,2048,0.018363555272420246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,2560,0.0362462235821618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,5120,0.04462933209207323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,1536,0.014850666125615438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,2048,0.016572443975342643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,2048,0.030301332473754883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,6144,0.07684889104631212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,1024,0.010915555887752108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,1536,0.013762666947311826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,768,0.008915555146005418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,1536,0.024683554967244465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,1024,0.012050666742854647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,512,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,3072,0.026016000244352553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,1024,0.018592000007629395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,768,0.011796444654464722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,256,0.005377777748637729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,768,0.01575377749072181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,512,0.01088177743885252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,128,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,256,0.010666666759385003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,3072,0.01984444426165687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,512,0.012999110751681857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,64,0.00462666650613149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,256,0.011071999867757162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,32,0.005154666801293691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,128,0.01053155544731352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,128,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,32,0.010050666828950247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,64,0.010370666782061258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,51200,0.2617937723795573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,65536,0.3741288979848226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,65536,0.2253511216905382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,16384,0.10223022434446548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,7168,0.08950755331251357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,12288,0.07359288798438178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,51200,0.1781680054134793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,16384,0.06300533480114408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,10240,0.06217600239647759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,12288,0.04751111070315043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,8192,0.05150755577617221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,16384,0.15604355600145128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,65536,0.6066231197781032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,12288,0.11763289239671494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,10240,0.045738667249679565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,7168,0.04309422108862135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,6144,0.03683111071586609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,8192,0.032920890384250216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,10240,0.09973066382937962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,5120,0.03023911184734768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,7168,0.027800000376171533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,8192,0.08068178097407024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,4096,0.024326221810446844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,7168,0.07122311327192518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,6144,0.061236441135406494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,5120,0.02141155633661482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,3584,0.0215013325214386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,3072,0.018757333358128864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,4096,0.017942221628295053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,5120,0.05159822106361389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,2560,0.016542222764756944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,3584,0.016893333858913846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,4096,0.042819556262758046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,2048,0.01311644415060679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,3072,0.015406222807036506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,3584,0.03826755616399977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,1536,0.011055111057228513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,2560,0.014153778553009033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,3072,0.033801777495278254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,1024,0.008664888640244802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,2048,0.012787555654843649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,2560,0.02897600001758999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,768,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,1536,0.01184177812602785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,1024,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,512,0.0058853332367208265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,768,0.010406222608354356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,1024,0.015507555670208402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,256,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,512,0.009666666388511658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,512,0.011027555498811932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,128,0.0041866666740841335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,256,0.009330666727489894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,6144,0.024489776955710515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,256,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,64,0.003879111260175705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,128,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,32,0.0041946665280395085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,128,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,64,0.008345777789751688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,32,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,2048,0.02471911079353756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,1536,0.02036088870631324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,65536,0.21806577841440836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,768,0.013222222526868185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,51200,0.24523644977145723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,16384,0.07262933254241943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,51200,0.17400266064537895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,16384,0.059176001283857554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,12288,0.06291644440756904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,51200,0.41594399346245664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,10240,0.04823199907938639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,12288,0.04698222213321262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,12288,0.1028604441218906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,8192,0.03891822364595201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,10240,0.041660444604025945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,10240,0.08656533559163411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,65536,0.31894222895304364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,7168,0.03401422169473436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,8192,0.03051911128891839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,6144,0.027660444378852844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,8192,0.06962844398286608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,51200,0.47489335801866317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,7168,0.06150755617353651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,5120,0.023481777972645227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,6144,0.023844444089465674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,65536,0.5223955578274196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,4096,0.020263999700546265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,6144,0.05367733372582329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,3584,0.01665155589580536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,5120,0.02068000038464864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,4096,0.017895110779338412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,16384,0.1356497738096449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,3584,0.016680889659457736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,3584,0.03349155518743727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,2560,0.013066666821638743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,3072,0.015393777026070489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,3072,0.029548443026012842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,2048,0.01089244418674045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,2560,0.014144000079896716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,2560,0.02572888963752323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,1536,0.008995555341243744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,2048,0.013059555656380124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,1024,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,2048,0.021703999903466966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,7168,0.02700622214211358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,1536,0.011763555308183035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,768,0.005889777921968036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,1536,0.017975111802419026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,1024,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,512,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,768,0.01035199976629681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,1024,0.013842666314707862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,256,0.0038977778620190094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,768,0.011766222616036734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,512,0.009708444277445475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,128,0.003530666646030214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,256,0.009364444348547194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,512,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,3072,0.014544000228246054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,64,0.003177777760558658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,32,0.0032293332947625052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,4096,0.03745955559942458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,128,0.009319999979601966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,256,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,128,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,64,0.007585778004593319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,65536,0.21988356113433838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,51200,0.1704648865593804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,32,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,16384,0.05183466606669956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,65536,0.17211643854777017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,12288,0.041532443629370794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,16384,0.04712444543838501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,51200,0.13667110602060953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,10240,0.03611822260750665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,12288,0.037795556916130915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,65536,0.4378017849392361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,16384,0.1144088904062907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,8192,0.025853332546022203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,12288,0.08688177665074666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,7168,0.022688888841205176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,10240,0.03394844465785556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,6144,0.019488889310095046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,8192,0.023717333873112995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,10240,0.07318133115768433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,7168,0.02309155629740821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,5120,0.015998222761683993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,8192,0.05905244747797648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,4096,0.013405332962671915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,7168,0.05204266640875074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,6144,0.018410666121376883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,5120,0.016514667206340365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,6144,0.045716444651285805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,5120,0.045273777511384755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,3584,0.01180355581972334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,3072,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,4096,0.014500444134076437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,5120,0.038732445902294584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,2560,0.009053332938088311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,4096,0.03207822309599982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,3584,0.01351288871632682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,2048,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,3072,0.012747555143303342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,3584,0.028672890530692205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,1536,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,3072,0.02535733415020837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,2560,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,1024,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,2560,0.022121777137120564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,1536,0.010666666759385003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,2048,0.018788443671332467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,768,0.004569777597983678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,1024,0.009847111172146266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,1536,0.015531554818153381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,512,0.0038693332009845306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,51200,0.34670045640733504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,1024,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,256,0.0031831111345026228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,768,0.009582222335868413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,128,0.0029066666546795103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,512,0.008999110923873054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,768,0.010407999985747868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,64,0.00282222218811512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,512,0.008972444468074376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,32,0.002883555574549569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,128,0.00850933293501536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,256,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,128,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,64,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,65536,0.1718240049150255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,32,0.006653333289755716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,51200,0.13711467054155138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,65536,0.17012977600097656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,16384,0.04411555661095513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,65536,0.3562355571322971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,51200,0.13571643829345703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,2048,0.011355555719799466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,12288,0.03375822305679321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,16384,0.04680711030960083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,51200,0.2813137902153863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,10240,0.02867822183503045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,8192,0.02141511109140184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,16384,0.09410844246546428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,12288,0.03771200113826328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,7168,0.014140443669425117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,12288,0.07161955700980292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,8192,0.02424088948302799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,6144,0.012166221936543783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,8192,0.04831199844678243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,256,0.00869599978129069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,7168,0.02166666587193807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,5120,0.010823110739390055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,7168,0.04281333420011732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,4096,0.008784888519181145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,6144,0.018567111757066514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,3584,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,6144,0.037512888511021934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,5120,0.03218844532966614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,4096,0.014766222900814481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,3072,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,4096,0.026480888326962788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,3584,0.013746666411558786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,2560,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,3584,0.02358755634890662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,3072,0.0131217779384719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,2048,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,3072,0.021106666988796655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,1536,0.00481599983241823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,2560,0.0184568895234002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,2048,0.011406222151385413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,1024,0.004210666649871402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,2048,0.01590755581855774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,1536,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,10240,0.03432888785998026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,768,0.003531555748648114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,1536,0.01344622257683012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,1024,0.009823111196359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,512,0.003154666680428717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,10240,0.059643553362952344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,1024,0.010533333652549319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,768,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,512,0.008943999807039896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,768,0.00904444439543618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,128,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,256,0.008486222061845992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,512,0.00795911086930169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,64,0.002287111141615444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,32,0.00254844439526399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,5120,0.01688088807794783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,128,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,128,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,65536,0.09903733597861396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,64,0.005843555761708154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,32,0.005670222143332164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,51200,0.07830133040746053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,2560,0.012102222277058495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,65536,0.17076178391774496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,16384,0.02981333269013299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,65536,0.31557512283325195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,12288,0.025791999366548326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,51200,0.136790222591824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,16384,0.04970755510860019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,16384,0.08306222491794162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,10240,0.016407110624843173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,12288,0.039817776944902204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,256,0.0027226667023367355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,12288,0.06314666403664483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,10240,0.03564888901180691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,7168,0.011752000285519494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,10240,0.05318489008479648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,8192,0.024587555064095393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,256,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,8192,0.04317511121431986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,7168,0.02166044380929735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,5120,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,7168,0.03791288865937127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,6144,0.019839111301634047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,4096,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,6144,0.033370667033725314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,5120,0.017916444275114272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,3584,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,5120,0.028597331709331934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,4096,0.014930667148696052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,3072,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,4096,0.023766222927305434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,3584,0.014793778459231058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,51200,0.24754489792717826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,3584,0.021611554755104914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,3072,0.013472889032628802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,2048,0.0047662220895290375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,3072,0.0191928893327713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,2560,0.012441777520709567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,1536,0.0042035554846127825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,2560,0.016915554801623028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,2048,0.01145155562294854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,1024,0.003552888830502828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,2048,0.014799111419253878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,6144,0.010320888625250923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,768,0.0031511110977994073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,1536,0.01239733315176434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,1024,0.009947555760542551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,512,0.00282222218811512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,1024,0.010049777726332346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,768,0.009588444398509132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,256,0.0025439999169773525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,512,0.008801778157552084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,768,0.008569777839713627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,512,0.007387555307812161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,256,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,64,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,256,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,128,0.008343111309740279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,32,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,128,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,2560,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,65536,0.08913689189487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,64,0.0053022222386466126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,51200,0.0718782212999132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,16384,0.02871733241611057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,32,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,8192,0.01330133361948861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,12288,0.02367288867632548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,10240,0.01444533301724328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,8192,0.01260355528857973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,65536,0.315282662709554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,16384,0.08303022384643555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,51200,0.24731289015875924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,1536,0.010747555229398938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,12288,0.06287822458479139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,7168,0.011212444139851464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,6144,0.009967111051082611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,10240,0.05305155449443393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,4096,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,5120,0.008643555144468943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,3584,0.007076444725195567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,8192,0.04309689005215963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,3072,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,2560,0.00545866663257281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,7168,0.03815022110939026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,6144,0.0329022208849589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,128,0.0022026666750510535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,2048,0.004567999806669024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,4096,0.023511999183230933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,1536,0.0041964443193541635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,1024,0.0034977777136696708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,3072,0.01898755629857381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,768,0.003208000005947219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,512,0.002842666581273079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,2048,0.014468444718254937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,1536,0.012207110722859701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,256,0.002452444492114915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,1024,0.009718221922715506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,768,0.008345777789751688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,128,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,512,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,256,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,32,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,128,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,64,0.005312888986534542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,65536,0.09090044763353135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,51200,0.07222844494713677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,32,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,16384,0.02637600070900387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,5120,0.02831377916865879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,3584,0.02125866711139679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,12288,0.019871110717455547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,10240,0.012190221912331052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,2560,0.016602666841612924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,51200,0.2469600041707357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,16384,0.08295110861460368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,8192,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,7168,0.009891555541091496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,6144,0.00870311094654931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,12288,0.06277777751286824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,64,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,8192,0.04282133446799385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,7168,0.037872887319988675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,4096,0.007370666497283512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,6144,0.03295288814438714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,3584,0.007038222418891058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,5120,0.008059555457697975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,3072,0.007704888780911763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,2560,0.006683555742104848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,5120,0.02808977829085456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,2048,0.004350222233268949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,4096,0.02346044447686937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,3584,0.02203377750184801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,1536,0.0038702223036024305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,3072,0.019723556107944913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,1024,0.0032017777363459268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,768,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,2560,0.017235555582576327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,512,0.002528888897763358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,2048,0.014827556080288358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,256,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,1536,0.012433778080675336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,1024,0.009947555760542551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,128,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,768,0.00847822262181176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,64,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,32,0.0018968888454967076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,512,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,256,0.005971555494599872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,128,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,64,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,10240,0.05288711190223694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,32,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,65536,0.31521333588494194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,16384,1.7010107040405273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,16384,5.264198303222656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,16384,6.1893971761067705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,16384,5.759176890055339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,12288,4.7380087110731335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,51200,7.270000881618923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,12288,1.3689795600043402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,12288,4.3639708624945746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,10240,1.103800031873915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,10240,3.5410063001844616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,12288,4.071384853786893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,10240,3.7163064744737415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,8192,0.8969679938422309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,10240,3.302100499471029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,8192,2.944463094075521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,8192,2.8670177459716797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,8192,2.6874987284342446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,7168,0.7532933553059896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,7168,2.927202648586697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,51200,16.78704155815972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,7168,2.630957285563151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,6144,1.9996888902452257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,7168,2.3338036007351346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,6144,0.6745528644985623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,6144,2.148729748196072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,5120,1.7163084877861872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,5120,0.5679226451449925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,5120,1.6841262181599934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,6144,2.0547858344184027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,4096,1.6245475345187719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,5120,1.6893164316813152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,4096,1.378584861755371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,4096,0.47405242919921875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,4096,1.3176292843288846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,3584,0.4205617904663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,3584,1.346387545267741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,3584,1.2910195456610787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,3072,1.0539404551188152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,3072,0.417980432510376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,3584,1.1752586364746094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,3072,1.0349644554985893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,2560,0.8717217975192599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,2560,0.31784534454345703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,3072,1.0101742214626737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,2048,0.6323039796617296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,2560,0.8904320398966471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,2560,0.8550160196092395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,2048,0.2682044506072998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,1536,0.4563075701395671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,2048,0.7380560239156088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,1536,0.21862488322787813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,2048,0.658715565999349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,1024,0.35485243797302246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,1536,0.5968880123562282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,1536,0.5166399743821886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,768,0.26554756694369847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,1024,0.17408443821801078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,512,0.2158808840645684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,51200,19.039647420247395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,1024,0.4413448969523112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,1024,0.38904889424641925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,768,0.32597332530551487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,256,0.13996711042192247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,512,0.1331004434161716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,768,0.32070933447943795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,512,0.34264800283643937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,128,0.11781688531239827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,256,0.2519235610961914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,256,0.20081688298119438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,512,0.29371288087632924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,128,0.11544622315300836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,64,0.09745955467224121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,128,0.24232265684339735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,32,0.0995288888613383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,128,0.14995822641584608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,768,0.1842097838719686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,64,0.2524648772345649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,32,0.24836354785495332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,256,0.11521777841779922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,51200,26.948320176866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,51200,5.595619625515408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,65536,7.457470364040798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,16384,5.013662974039714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,51200,13.198321872287325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,16384,5.489590115017361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,16384,1.397613313462999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,65536,16.937330457899304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,12288,3.8960045708550344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,16384,4.484793768988715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,12288,3.614716423882378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,12288,1.0128569073147244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,12288,3.1742026011149087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,51200,16.92354498969184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,51200,21.683178371853298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,10240,2.8150586022271047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,10240,3.229467603895399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,10240,0.8295973141988119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,10240,2.6019679175482855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,8192,0.6902373101976184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,8192,2.4813946617974176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,8192,2.320624033610026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,7168,1.9329679277208116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,7168,0.6149155298868815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,8192,2.136367162068685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,7168,1.9639502631293402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,6144,1.6030747095743816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,6144,0.5314417945014106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,7168,1.8402603997124567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,6144,1.671916431850857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,5120,0.4470213254292806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,5120,1.5785253312852647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,6144,1.5418035719129775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,5120,1.3819093704223633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,4096,1.0920346577962239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,4096,0.36474932564629453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,5120,1.2955857382880318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,4096,1.091692394680447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,3584,0.32928087976243764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,3584,0.9085280100504557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,4096,1.0530160268147786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,3584,1.0050079557630751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,65536,29.702528211805557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,3072,0.7972782452901205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,3072,0.2870604462093777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,3584,0.9066978030734592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,3072,0.8206808831956652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,2560,0.6854035589430066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,2560,0.2532586786482069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,3072,0.7530710962083605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,2048,0.5675582355923122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,2560,0.6895395384894477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,2048,0.2191386620203654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,1536,0.38816266589694554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,2048,0.5607360204060873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,2560,0.6568542056613499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,1536,0.17219111654493544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,1024,0.28992001215616864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,2048,0.5316550996568468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,1024,0.13951288329230413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,768,0.23338132434421113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,1536,0.49178489049275714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,1024,0.3313475449879964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,1536,0.40540801154242617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,512,0.170960889922248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,1024,0.30766132142808705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,768,0.14167466428544787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,768,0.27335556348164874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,256,0.133586671617296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,512,0.22078222698635527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,512,0.10759822527567546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,768,0.25061244434780544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,128,0.09249511029985215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,256,0.20005867216322157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,512,0.2059253321753608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,64,0.07809599902894762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,128,0.18931110699971518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,128,0.09366399712032741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,32,0.07856533262464735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,128,0.11912088923984104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,256,0.1808462275399102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,64,0.1919564406077067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,32,0.19576444890764025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,256,0.09215822484758164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,65536,2.2371893988715277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,51200,4.415538787841797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,51200,1.7161253823174372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,65536,6.078169928656684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,16384,1.6241644753350153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,16384,0.43843910429212785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,51200,4.920270284016927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,16384,1.7940444946289062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,51200,7.053056081136067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,12288,0.3310488859812419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,65536,6.334245469835069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,16384,1.4670675065782335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,12288,1.2163466347588432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,12288,1.164903958638509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,12288,1.0401617685953777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,65536,8.882634480794271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,10240,0.9809360504150391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,10240,0.3165671030680339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,8192,0.2937733332316081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,10240,0.9699857499864367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,8192,0.8053341971503364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,7168,0.6225920253329807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,10240,0.8315297762552897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,8192,0.7564284536573621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,7168,0.24723643726772734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,8192,0.6300355593363444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,6144,0.639087994893392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,7168,0.6509671211242676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,6144,0.2199288871553209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,7168,0.5492124557495117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,6144,0.5704551272922093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,5120,0.42066134346856016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,5120,0.20057866308424208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,6144,0.4788524309794108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,4096,0.3351653416951497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,4096,0.1305262247721354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,4096,0.35333066516452366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,5120,0.44940710067749023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,3584,0.3012746704949273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,5120,0.39781689643859863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,3584,0.11472354994879828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,4096,0.3630337715148926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,3584,0.4406782256232367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,2560,0.21065066920386422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,3072,0.10160711076524521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,3072,0.3347333272298177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,3584,0.29400266541375053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,3072,0.2577759954664442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,2560,0.08904355764389038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,2560,0.2348719967736138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,2048,0.07512178023656209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,1536,0.14055911699930826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,2048,0.19094044632381865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,2560,0.21597245004442003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,1024,0.09056799941592747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,1536,0.06175644530190361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,2048,0.17740889390309653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,1536,0.14656710624694824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,3072,0.25313599904378253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,768,0.06893422206242879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,1536,0.14270667235056558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,1024,0.05014400018586052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,1024,0.11361866527133518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,512,0.05168799890412224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,768,0.0457839998934004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,768,0.0884737769762675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,1024,0.10901421970791286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,256,0.035572446054882474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,2048,0.1703324450386895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,512,0.07582488987180921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,512,0.041803555356131665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,768,0.09047555261188084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,128,0.02850133180618286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,256,0.03727466530270047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,512,0.07515822516547309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,256,0.06810044580035739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,128,0.03717333409521315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,32,0.02607911162906223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,128,0.06445866823196411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,256,0.05858133236567179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,128,0.04457599918047587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,64,0.065482669406467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,32,0.06673777765697904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,64,0.02611555490228865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,65536,1.5356346766153972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,51200,3.254585689968533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,51200,1.2243884404500325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,65536,4.258783128526476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,16384,1.0800000296698677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,16384,0.33362311787075466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,16384,1.321423954433865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,16384,1.1459004084269206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,65536,4.772478315565321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,51200,5.061921861436632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,12288,0.29883376757303876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,12288,0.8315528763665093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,51200,3.7915882534450955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,12288,0.7884977658589681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,65536,7.593291388617621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,10240,0.24462223052978516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,12288,0.8939849005805122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,10240,0.7798346413506402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,8192,0.6264257960849339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,10240,0.6067324214511448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,8192,0.17932532893286812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,8192,0.5622924698723687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,7168,0.1595031155480279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,7168,0.48652442296346027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,10240,0.7575484381781684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,6144,0.39206756485833066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,8192,0.4881422254774306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,6144,0.1389075517654419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,7168,0.43111112382676864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,7168,0.7209759818183051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,5120,0.3509848912556966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,6144,0.40190400017632383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,5120,0.11729333135816787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,5120,0.3493662145402696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,4096,0.3350631131066217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,4096,0.09731911288367377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,3584,0.23198310534159342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,4096,0.32527022891574436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,5120,0.31084977255927193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,3584,0.09094844261805217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,4096,0.25460889604356557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,3584,0.24740444289313424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,3072,0.19573688507080078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,2560,0.15147199895646837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,3072,0.07846399810579088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,3072,0.21186221970452201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,3584,0.22422843509250215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,2560,0.06873422198825412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,2048,0.14257600572374132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,2560,0.20508178075154623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,3072,0.1978480021158854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,1536,0.1040266686015659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,2048,0.0580622222688463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,2048,0.14315466086069742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,6144,0.37751022974650067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,2560,0.1731413337919447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,1024,0.07665955358081393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,1536,0.04777600036727058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,1536,0.11497422059377034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,768,0.06138133340411716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,2048,0.14168000221252441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,1024,0.08625955714119805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,1536,0.11382488409678142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,512,0.04388622111744351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,768,0.06940888696246676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,768,0.037456889947255455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,1024,0.08707822031444973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,256,0.030031998952229817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,512,0.034254223108291626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,768,0.07285066445668538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,128,0.022829333941141765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,256,0.05289511217011345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,256,0.030616889397303265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,512,0.05960710843404134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,64,0.020092444287406076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,256,0.046801779005262584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,128,0.030662222041024104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,32,0.020548444655206468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,128,0.05006044440799289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,128,0.03610400027698941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,32,0.051593777206208974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,1024,0.040339556005265974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,512,0.058986663818359375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,64,0.05061155557632446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,65536,1.3527741962009008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,51200,3.0966977013481984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,65536,3.459910286797417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,51200,1.0407653384738498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,16384,0.9116107092963325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,16384,1.005981339348687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,16384,0.28396532270643443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,16384,0.9301492902967664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,65536,5.343128204345703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,65536,3.989178551567925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,12288,0.6097680197821723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,51200,4.606633928087023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,51200,3.211751937866211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,12288,0.22502043512132433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,12288,0.7143191231621636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,10240,0.5933715502421061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,12288,0.6644657982720269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,10240,0.1857066684299045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,10240,0.6247484419080946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,8192,0.4402444362640381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,8192,0.1507564385732015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,10240,0.5484399795532227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,7168,0.3457635508643256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,8192,0.5307031207614475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,7168,0.1331013308631049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,8192,0.42148534456888836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,6144,0.30957423316107857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,7168,0.5045066409640843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,6144,0.11553777588738336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,7168,0.36434221267700195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,5120,0.28077154689364964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,6144,0.3412177827623155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,5120,0.10166311264038086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,5120,0.28689689106411403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,4096,0.22079022725423178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,6144,0.3183635605706109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,4096,0.08348266945944892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,4096,0.2258791128794352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,5120,0.2655155658721924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,3584,0.20516533321804473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,3072,0.1639502180947198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,4096,0.2161404424243503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,2560,0.14703467157151964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,3072,0.06605866882536146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,3584,0.19112088945176867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,3072,0.17314488357967803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,2048,0.10911289188596938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,2560,0.058430221345689565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,3072,0.17085866133371988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,2560,0.1477137804031372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,1536,0.08324978086683485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,2048,0.04993866549597847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,2560,0.14460710684458414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,2048,0.12195377879672581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,3584,0.18813244501749674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,1536,0.0417902237839169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,1024,0.06719999843173556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,2048,0.12016177177429199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,1536,0.0974666674931844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,768,0.04856444398562113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,3584,0.07633599970075819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,1024,0.03569688730769687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,1536,0.09847288661532932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,1024,0.07429777913623385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,768,0.032905777295430504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,768,0.059213333659701876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,1024,0.07651378048790826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,256,0.030343112018373277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,512,0.030365334616767034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,512,0.05038933290375603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,768,0.0633537769317627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,128,0.0214035560687383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,256,0.027098667290475633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,512,0.053048001395331494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,256,0.04529777831501431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,64,0.01787644459141625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,32,0.01792888840039571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,128,0.042556444803873696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,128,0.02717688845263587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,256,0.04131999942991469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,128,0.03146933184729682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,64,0.04335111048486498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,32,0.04409689042303297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,512,0.037552889850404524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,65536,1.3316044277615018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,65536,18.259989420572918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,51200,2.3824310302734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,16384,0.8328373167249891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,65536,3.11297247144911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,51200,1.0574827194213867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,16384,0.27104443973965114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,51200,3.3681697845458984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,16384,0.8388080067104764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,51200,2.5793955061170792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,12288,0.7181590927971734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,12288,0.19085333082411024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,16384,0.7502017550998263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,65536,3.283545811971029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,10240,0.48246224721272785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,12288,0.5839049021402994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,65536,4.392764621310763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,10240,0.16711200608147514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,12288,0.5199368794759115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,8192,0.3844826751285129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,8192,0.12662933932410345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,10240,0.5045822461446127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,7168,0.3264319896697998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,10240,0.42274755901760525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,7168,0.1134106715520223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,8192,0.4637004534403483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,6144,0.26216088400946724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,8192,0.33881423208448624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,7168,0.34099555015563965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,5120,0.21428977118598092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,6144,0.09992177618874444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,6144,0.30221332444085014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,7168,0.33384354909261066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,4096,0.17435022195180258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,5120,0.24071200688680014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,6144,0.260079993142022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,5120,0.0848826633559333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,3584,0.13364622328016493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,4096,0.07464088996251424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,4096,0.18780267238616943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,3072,0.14145955774519178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,5120,0.2181839942932129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,3584,0.06341599755816989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,3584,0.1655635568830702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,4096,0.18243111504448783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,2560,0.1094942225350274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,3072,0.056840889983707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,3584,0.15973333517710367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,2048,0.10068266921573216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,2560,0.05006133185492622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,2560,0.1238293382856581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,3072,0.14245600170559353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,1536,0.0660382244322035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,2048,0.04275644487804837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,2048,0.100582222143809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,2560,0.12026845084296332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,1024,0.05096888873312208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,1536,0.035880890157487653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,1536,0.0798471106423272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,2048,0.1025902231534322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,768,0.0406951109568278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,1024,0.030060443613264296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,1024,0.06119822131262886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,1536,0.08294222089979383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,512,0.02958577871322632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,768,0.026695999834272597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,768,0.049397332800759204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,1024,0.06339110930760701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,512,0.02437866727511088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,512,0.04126666651831733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,768,0.053929779264662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,128,0.015391111373901367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,256,0.02275466587808397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,256,0.03642222285270691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,64,0.01331555595000585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,128,0.021394666698243883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,128,0.03413155674934387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,256,0.03469689024819268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,3072,0.1408862272898356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,128,0.027433777848879497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,64,0.03426400158140395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,32,0.03501333461867438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,256,0.019632889164818656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,512,0.04480266571044922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,32,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,65536,1.088897811041938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,51200,0.7968977822197808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,51200,2.1521288553873696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,16384,0.892905765109592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,16384,0.2509964572058784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,65536,3.057534323798286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,16384,0.7539155748155383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,16384,0.6277742385864258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,12288,0.5841395590040419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,65536,2.883456971910265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,51200,3.040918138292101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,12288,0.5519030888875326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,51200,2.2585315704345703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,65536,3.9279946221245656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,10240,0.3875288963317871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,10240,0.1419395605723063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,12288,0.46150578392876523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,8192,0.33286934428744847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,8192,0.11346044805314805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,10240,0.4418044355180528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,8192,0.3433395491706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,7168,0.2747635576460096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,10240,0.3782293266720242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,7168,0.10030133194393581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,6144,0.2391200065612793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,8192,0.3136906623840332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,6144,0.08867377705044216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,5120,0.20280088318718803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,6144,0.3088471094767253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,12288,0.16501243909200033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,7168,0.26589155197143555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,4096,0.14787644810146755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,5120,0.20906933148701987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,3584,0.13626933097839355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,4096,0.16394400596618652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,5120,0.19483466943105063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,4096,0.1595662169986301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,3584,0.05712799893485176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,3072,0.12433066633012559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,3584,0.14471288522084555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,3584,0.14325066407521567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,7168,0.40132445759243435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,3072,0.05126844512091743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,2560,0.09915644592709011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,3072,0.12591289149390325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,3072,0.1251493295033773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,2560,0.04447022080421448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,2560,0.10902666383319431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,5120,0.08524800009197658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,2048,0.08305866850747003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,1536,0.0669671098391215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,2048,0.03823022378815545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,4096,0.06700177987416585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,2048,0.08981421920988296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,1024,0.04328711165322197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,1536,0.031839112440745033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,1536,0.07107555866241455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,2048,0.09228355354732937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,768,0.03252977795071072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,1024,0.026693332526418898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,1024,0.054281777805752225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,1536,0.07481155792872111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,512,0.025008888708220586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,768,0.04407644602987501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,1024,0.057598219977484814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,512,0.02198133369286855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,768,0.0483511123392317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,512,0.03543466660711501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,256,0.017680888374646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,6144,0.23159466849433052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,128,0.014137778017255994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,256,0.019966221517986722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,2560,0.10881155729293823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,256,0.031001776456832886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,64,0.013068444199032254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,128,0.029488888051774766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,256,0.03223289052645365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,128,0.01963200006220076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,32,0.013452444639470843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,64,0.029161777761247423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,128,0.02420266634888119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,32,0.029863999949561224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,768,0.024533333049880132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,512,0.04087555408477783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,65536,0.910455067952474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,51200,0.6929795477125379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,51200,1.7645457585652669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,65536,2.1626400417751737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,16384,0.7025778028700086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,16384,0.19806222120920816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,12288,0.4231848981645372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,16384,0.659921752081977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,51200,2.5552967919243708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,51200,2.009895112779405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,16384,0.5879013273451064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,12288,0.451706674363878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,65536,3.550639046563042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,65536,2.576742172241211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,10240,0.1232266690995958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,12288,0.4348017904493544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,10240,0.38868623309665257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,8192,0.27347199122111004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,8192,0.09837600257661606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,10240,0.35071465704176163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,7168,0.25675733884175617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,8192,0.3528880013359918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,7168,0.09081244468688965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,8192,0.27863823042975533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,7168,0.28451556629604763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,12288,0.1453582180870904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,5120,0.1587280035018921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,6144,0.07776977618535359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,6144,0.21876088778177896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,10240,0.36092978053622776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,5120,0.06709777646594577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,6144,0.21238577365875244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,5120,0.18185599644978842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,4096,0.14443555143144396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,3584,0.11559911568959554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,4096,0.1482035583919949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,3584,0.05159733361668057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,6144,0.2045066621568468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,3072,0.10022488567564224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,3584,0.1326871050728692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,7168,0.24946488274468315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,3072,0.11059200101428562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,2560,0.08931733502282037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,4096,0.12854577435387507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,3072,0.04510133465131124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,2048,0.0655226641231113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,5120,0.18049777878655326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,2560,0.03992444607946608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,3072,0.11618399620056152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,1536,0.05356711149215698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,2560,0.09556888871722752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,4096,0.057687997817993164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,2048,0.034381333324644305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,2560,0.099199997054206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,1024,0.03855822152561612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,1536,0.02865866820017497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,1536,0.0626595550113254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,2048,0.08509688907199436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,768,0.028993778758578833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,1024,0.02390755547417535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,1536,0.06836977932188246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,1024,0.053047110637029014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,768,0.038904001315434776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,768,0.022073777185546026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,512,0.020823111136754353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,512,0.019595555133289762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,768,0.044682668315039746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,256,0.015408888459205627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,512,0.03137600090768602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,512,0.03791466686460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,256,0.018001778258217704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,256,0.027013333307372198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,2048,0.07947644260194567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,256,0.028024001253975764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,64,0.010718222293588849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,128,0.025751999682850305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,128,0.017824000782436795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,32,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,3584,0.12689067257775202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,1024,0.04768266611629062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,128,0.021635555558734473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,64,0.025762667258580525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,32,0.025760889053344727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,128,0.01239466667175293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,65536,0.7592195404900445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,51200,1.731877326965332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,16384,0.41147732734680176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,51200,0.6276337835523818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,16384,0.19607199562920463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,51200,2.227277331882053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,65536,2.720615175035265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,16384,0.6804391013251411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,16384,0.49400001102023655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,65536,2.2401385837131076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,51200,1.6863262388441298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,12288,0.1263724433051215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,12288,0.38103556632995605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,10240,0.24899199273851183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,65536,2.03547117445204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,8192,0.20061333974202475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,10240,0.10788355271021526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,10240,0.32826044824388295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,12288,0.37463998794555664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,8192,0.08903021944893731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,7168,0.22366044256422254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,8192,0.28722045156690806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,10240,0.31879909833272296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,6144,0.17575555377536348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,7168,0.0780186653137207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,8192,0.24289155006408691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,7168,0.21441511313120523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,5120,0.14929955535464817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,12288,0.39417065514458555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,6144,0.08560978041754828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,6144,0.188664886686537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,4096,0.11555733945634629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,7168,0.21476711167229545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,5120,0.05921866496404012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,5120,0.154950221379598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,3584,0.10290844572914971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,6144,0.1914800008138021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,4096,0.049485332436031766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,4096,0.14139733049604628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,3072,0.08196266492207845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,3584,0.04508444335725573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,3584,0.1102364460627238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,4096,0.1303102175394694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,2560,0.07518221934636433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,3072,0.03954044315550063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,3072,0.09548533625072902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,2048,0.05702222055859036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,2560,0.03523644473817613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,2560,0.08226933081944783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,3072,0.10178311003579034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,1536,0.04164888792567783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,2048,0.029551111989551123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,2048,0.06784266895718045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,1024,0.030545777744717065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,2560,0.08842311302820842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,1536,0.024862221545643274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,2048,0.0740764472219679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,1536,0.05389066537221273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,768,0.02424266603257921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,1024,0.021387555532985266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,1024,0.04081599911053976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,1536,0.06133421924379137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,768,0.01920000049802992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,1024,0.047284444173177086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,5120,0.15760800573560926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,256,0.012925333446926542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,512,0.017464000317785475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,768,0.040341334210501775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,512,0.027507555153634813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,128,0.013510222236315409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,3584,0.11763377984364827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,256,0.01663733356528812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,256,0.023357333408461675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,512,0.032784889141718544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,64,0.012367111113336353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,128,0.016498666670587327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,128,0.02237777743074629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,256,0.025285333395004272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,32,0.011709333293967776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,64,0.02238399949338701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,128,0.019552888141738046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,32,0.022240888741281297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,512,0.016713778177897137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,768,0.0336151123046875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,65536,0.5857120090060765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,65536,1.3433208465576172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,51200,1.0285973019070096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,51200,0.4661617808871799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,16384,0.36578043301900226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,16384,0.4294897715250651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,16384,0.14328622817993164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,51200,1.6188880072699652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,65536,2.271488825480143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,12288,0.28592443466186523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,51200,1.4178027047051325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,65536,1.8487101660834417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,16384,0.4382755491468642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,12288,0.10691910982131958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,10240,0.24280089802212187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,12288,0.3130231168535021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,8192,0.19017510943942598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,8192,0.07268000311321683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,7168,0.16427466604444715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,12288,0.31559377246432835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,8192,0.1989875502056546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,10240,0.2640106678009033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,6144,0.14440888828701443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,8192,0.21902667151557076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,7168,0.06634666522343953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,7168,0.17881245083279082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,5120,0.11525244183010525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,6144,0.0579928888214959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,6144,0.1551253398259481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,7168,0.1965644492043389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,4096,0.09431466791364883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,5120,0.04997244477272034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,5120,0.1292168829176161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,6144,0.15840266810523138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,3584,0.0852853324678209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,4096,0.041215999258889094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,4096,0.11127821604410808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,10240,0.10070844491322835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,3072,0.06650933292176989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,3584,0.03704799877272712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,10240,0.2807199954986572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,3584,0.09186133411195542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,4096,0.11042844586902195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,2560,0.059228446748521596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,3072,0.03272711237271627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,3584,0.09940799739625718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,3072,0.08078489038679335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,2048,0.045180443260404796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,2560,0.02931200133429633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,1536,0.03511555658446418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,3072,0.0883351100815667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,2048,0.025444444682863023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,2560,0.07528977923923068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,1024,0.02516355613867442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,1536,0.02180444366402096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,1536,0.045920888582865395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,2048,0.06531377633412679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,768,0.018908444378111098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,1024,0.01847111185391744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,1536,0.05315999852286445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,1024,0.034252444903055825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,512,0.014255111416180929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,5120,0.13548889425065783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,768,0.028390222125583228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,768,0.017272000511487324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,1024,0.04102311200565762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,256,0.010695111420419482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,512,0.015386665860811869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,512,0.022639110684394836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,768,0.034225778447257146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,128,0.009287999735938178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,256,0.019888889458444383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,256,0.014463111758232117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,64,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,512,0.028223110569847956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,2560,0.06855733527077569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,32,0.00851288851764467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,256,0.022336888644430373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,128,0.018971555762820773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,2048,0.05710311068428887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,64,0.018562666243977018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,128,0.017567111386193168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,32,0.01867733399073283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,65536,0.5329270892673069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,51200,0.9183502197265625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,128,0.014869333969222175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,65536,1.4649982452392578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,16384,0.3468426598442926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,51200,1.514463954501682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,16384,0.1384728882047865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,51200,1.2537733713785808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,65536,1.9967767927381728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,65536,1.654929796854655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,12288,0.2703084415859646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,16384,0.38755022154914004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,12288,0.10729510916603936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,10240,0.22008889251285127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,12288,0.2775413460201687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,8192,0.1740337742699517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,10240,0.09501066472795273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,10240,0.2258791128794352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,12288,0.2843200100792779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,51200,0.4380168914794922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,7168,0.15961066881815592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,8192,0.06735910971959432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,10240,0.24398933516608345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,8192,0.20068532890743682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,16384,0.35700355635748965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,6144,0.1302373276816474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,7168,0.06014044417275322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,7168,0.15986843903859457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,8192,0.19305955039130318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,5120,0.10698133044772679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,6144,0.052503999736573964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,6144,0.14038577344682482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,4096,0.08236533403396606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,7168,0.18288089169396293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,5120,0.04453600115246243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,5120,0.11588799953460693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,3584,0.07770578066507976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,4096,0.03687644335958693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,6144,0.14502488242255315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,3072,0.0624613364537557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,4096,0.09355911281373765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,3584,0.08318310976028442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,2560,0.053103999959097974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,3072,0.029436445898479883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,3072,0.07195288605160184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,3584,0.09133866760465835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,2048,0.04455822375085619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,2560,0.026641777820057336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,3072,0.0815253323978848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,2048,0.023183999790085688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,2048,0.05153333478503757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,1536,0.033929778469933405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,2048,0.0591813325881958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,1536,0.04133955637613932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,1024,0.02274133265018463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,1536,0.019675556156370375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,5120,0.12279378043280707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,1536,0.048878222703933716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,768,0.018720888429217868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,4096,0.10364622539944118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,1024,0.030913776821560327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,768,0.016065778003798593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,768,0.02584533393383026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,512,0.01534844438234965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,2560,0.061405334207746715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,768,0.0318159990840488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,2560,0.06974933544794719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,256,0.011389333340856763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,512,0.026512000295850966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,256,0.01789777808719211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,128,0.009088888764381409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,256,0.013759111364682516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,256,0.020294222566816542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,3584,0.033420443534851074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,64,0.00793333351612091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,128,0.013600000076823764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,1024,0.01651022169325087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,32,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,128,0.017110221915774874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,64,0.016890666551060148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,1024,0.037498666180504694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,32,0.016935999194780987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,512,0.014470222923490735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,512,0.020555555820465088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,65536,0.4626915719774034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,65536,0.9433795081244575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,51200,0.8809066878424751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,16384,0.27951023313734263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,51200,0.36276178889804417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,16384,0.11523644129435222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,65536,1.5993386374579535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,51200,1.1486675474378798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,16384,0.4191982216305203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,12288,0.2368933359781901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,65536,1.46142397986518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,51200,1.1232817967732747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,12288,0.09205866522259182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,10240,0.186880005730523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,16384,0.397216002146403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,12288,0.27849067582024467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,8192,0.1483128865559896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,10240,0.07427733474307589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,10240,0.21380533112419975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,128,0.01752444439464145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,12288,0.2572444544898139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,7168,0.12544443872239855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,8192,0.07303644551171197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,8192,0.1690373288260566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,6144,0.10383288727866279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,10240,0.21722400188446045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,7168,0.053631110323799976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,5120,0.09255111217498779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,6144,0.04775911238458422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,7168,0.14817244476742214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,8192,0.17316977183024088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,6144,0.1533715592490302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,4096,0.0699253347184923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,7168,0.15187377399868435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,5120,0.040087110466427274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,3584,0.05859111415015327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,6144,0.14832177427079943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,5120,0.10277777910232544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,4096,0.032487998406092324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,4096,0.08577244149314033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,3072,0.05158488949139913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,5120,0.1138391097386678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,3584,0.02938666608598497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,2560,0.0424124432934655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,4096,0.09282133314344619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,3072,0.027007111244731482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,3072,0.06455555227067736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,3584,0.08180355363421969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,2048,0.036385777923795916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,2560,0.024340444140964087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,3072,0.07329066594441731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,1536,0.02770311137040456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,2048,0.021132444341977436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,2560,0.06397333410051134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,2048,0.04656799965434604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,1024,0.019534221953815885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,1536,0.03754933343993293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,2048,0.05407289001676771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,1024,0.02770044406255086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,1536,0.044440888696246676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,1024,0.016223111086421542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,768,0.014498665928840637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,768,0.01478666729397244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,3584,0.07487555344899495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,512,0.01129155523247189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,768,0.029495110114415485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,512,0.013532444834709167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,256,0.008618666893906063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,512,0.01870133313867781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,2560,0.055204444461398654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,128,0.007368000017272101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,256,0.012800888882742988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,256,0.016192000773217943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,512,0.023798222343126934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,64,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,1536,0.018432888719770644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,128,0.012876444392734103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,32,0.0075626663035816616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,128,0.015683554940753512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,128,0.016435555285877652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,32,0.015141333142916361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,64,0.015248888068728976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,1024,0.03404888841840956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,768,0.023403555154800415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,65536,0.4165964391496446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,65536,0.8891306983100044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,256,0.019828443725903828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,16384,0.25495023197597927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,51200,0.7943511009216309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,51200,0.32909954918755424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,51200,0.9583777321709527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,65536,1.2370079888237846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,12288,0.18795377678341338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,16384,0.27092356152004665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,65536,1.279062271118164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,51200,0.9559439553154839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,12288,0.08061689138412476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,10240,0.1705155505074395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,16384,0.30513866742451984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,12288,0.2222835487789578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,8192,0.13438933425479466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,10240,0.0684373312526279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,10240,0.19200532966189912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,7168,0.11334933174981011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,8192,0.05480088790257772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,8192,0.14239554935031468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,10240,0.19458756181928846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,6144,0.09893688890669082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,7168,0.04831288920508491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,8192,0.15396622816721597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,7168,0.12443644470638698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,5120,0.08238399691051908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,6144,0.04086577892303467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,16384,0.11481689082251655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,6144,0.10997955666648017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,4096,0.06037511428197225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,5120,0.034509334299299456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,6144,0.11842666731940375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,3584,0.05584088961283366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,4096,0.02870222263866001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,5120,0.10041333569420709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,4096,0.08243555492824979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,12288,0.23023645083109537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,3072,0.04623288909594218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,3584,0.06486311223771837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,3584,0.07407822211583455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,3072,0.05651999844445122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,3072,0.02360444433159298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,2560,0.042577776643964976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,3072,0.06617244746949938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,7168,0.13719111018710667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,2048,0.033795555432637535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,2560,0.021407110823525324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,5120,0.0916444460550944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,2560,0.04871199859513176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,1536,0.02607022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,4096,0.07364533344904582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,2048,0.01903377804491255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,2048,0.04109866751564874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,1024,0.01687022215790219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,1536,0.01682577696111467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,1536,0.032955557107925415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,2048,0.04902044600910611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,3584,0.026196445027987163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,768,0.013342222405804528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,1024,0.014492443866199918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,1536,0.039862222141689725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,512,0.010994666980372535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,768,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,768,0.020957332518365648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,1024,0.030971556901931763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,256,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,768,0.026272889640596177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,512,0.012728888955381183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,256,0.011988444460762871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,512,0.02196088929971059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,2560,0.05766044722663032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,256,0.019507555498017203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,128,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,128,0.011951111257076263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,64,0.005709333138333426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,32,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,1024,0.024658666716681585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,128,0.014829332629839579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,64,0.01348000019788742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,32,0.013479111095269522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,512,0.016748444901572335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,256,0.014498665928840637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,128,0.006737777756320105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,65536,0.3353671232859294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,16384,0.18125511540306938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,51200,0.2683848804897732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,65536,0.9655315611097547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,16384,0.2523039976755778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,16384,0.0860053300857544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,65536,1.0180497699313693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,12288,0.13775022824605307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,51200,0.7902533213297526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,16384,0.25677066379123265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,12288,0.0677973363134596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,65536,0.5948693487379286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,10240,0.12344355053371853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,12288,0.17830044693417021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,8192,0.09487999810112847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,10240,0.05777155690722995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,51200,0.4893804656134711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,12288,0.20083999633789062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,10240,0.1479475498199463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,7168,0.0829262203640408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,8192,0.0454968876308865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,8192,0.11905154916975234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,10240,0.16544977823893228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,6144,0.07868178023232354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,7168,0.04081688986884223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,7168,0.10805333322948879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,8192,0.1363511085510254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,5120,0.05973066886266073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,6144,0.034908443689346313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,6144,0.09046044614579941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,4096,0.043475555049048535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,51200,0.7782542440626355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,5120,0.030240890052583482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,3584,0.03835733400450812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,5120,0.07680888970692952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,4096,0.02531733281082577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,4096,0.06164266665776571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,5120,0.0912604464424981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,3584,0.022963555322753057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,4096,0.0746213330162896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,3584,0.055439111259248525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,2560,0.027930667002995808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,3072,0.021194666624069214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,3584,0.06628621949089898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,2560,0.04234933190875583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,3072,0.058000889089372426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,2560,0.01926666663752662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,2048,0.023447111248970032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,2048,0.01720266706413693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,2560,0.05162488751941257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,1536,0.01923199991385142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,7168,0.1204648945066664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,2048,0.04387911160786947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,1536,0.015492444237073263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,1024,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,1536,0.028696888022952612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,3072,0.037075555986828275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,1536,0.03610400027698941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,768,0.011218667030334473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,1024,0.02161955502298143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,1024,0.013412444127930535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,3072,0.04877688818507724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,1024,0.028210666444566514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,512,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,768,0.012448000411192576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,768,0.018230222993426852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,256,0.006265777680608962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,512,0.011743110915025076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,768,0.02369511127471924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,512,0.014828445182906257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,128,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,256,0.012660444610648684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,256,0.011034666664070554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,2048,0.035551998350355364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,512,0.020023110840055678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,64,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,32,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,6144,0.1039395531018575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,256,0.01626844373014238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,64,0.011752888560295105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,128,0.013423110875818463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,32,0.011750222080283694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,65536,0.5720986790127224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,65536,0.27525422308180064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,51200,0.4001315434773763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,65536,0.8021821975708008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,51200,0.21974489423963758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,128,0.011055111057228513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,16384,0.1464266644583808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,128,0.012112889024946423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,65536,0.8528266482883029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,12288,0.09711288743548924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,16384,0.07425689034991793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,51200,0.6566586494445801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,16384,0.19747466511196562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,10240,0.09961955414877997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,12288,0.06221422221925524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,51200,0.6643502447340224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,8192,0.07407822211583455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,10240,0.05761155817243788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,16384,0.2207875516679552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,10240,0.126310216055976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,7168,0.06734844711091784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,12288,0.17298844125535753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,8192,0.04322488771544563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,8192,0.10581600003772312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,10240,0.14335821734534368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,6144,0.059005333317650684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,7168,0.03774844606717428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,7168,0.0902328888575236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,8192,0.11805955568949382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,5120,0.04814933405982124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,6144,0.03229511115286086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,4096,0.03475022315979004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,6144,0.07662577761544122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,7168,0.10428888930214776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,5120,0.027673777606752183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,3584,0.029876444074842665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,5120,0.06438844733768039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,6144,0.08950666586558025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,4096,0.023204444183243647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,3072,0.02552177839808994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,4096,0.05304888884226481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,5120,0.07654400004280938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,3584,0.020799110333124798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,2560,0.021772444248199463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,4096,0.06466044320000543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,3072,0.018644443816608854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,3072,0.04190399911668566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,3584,0.05734044313430786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,2048,0.018385777870814007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,2560,0.0181342214345932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,2560,0.03611288799179925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,3072,0.05144977900716993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,1536,0.014612444572978549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,2048,0.01643733349111345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,2560,0.04486755530039469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,2048,0.03044355577892727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,1024,0.010768888725174798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,1536,0.013744889034165276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,1536,0.024503111839294434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,768,0.009029332962301042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,1024,0.012071111136012606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,1024,0.018618666463428073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,1536,0.030208888981077407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,512,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,1024,0.02416177756256527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,768,0.015592000550693937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,3584,0.04764622118737963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,512,0.012977778083748288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,12288,0.14619021945529515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,256,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,768,0.02109066645304362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,128,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,256,0.010291555689440833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,512,0.018920888503392536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,256,0.011011555790901184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,128,0.010346666806273991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,64,0.004520888957712385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,256,0.015828443898095023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,128,0.0107342220014996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,32,0.004563555535342959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,2048,0.03768444392416213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,128,0.012311110893885294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,64,0.010048888623714447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,32,0.010048000348938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,65536,0.3894160058763292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,768,0.011391999820868174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,51200,0.291793770260281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,512,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,16384,0.10547733306884766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,65536,0.6056257883707682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,51200,0.17779821819729277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,51200,0.6306515799628364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,16384,0.0606479975912306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,12288,0.07065511412090726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,65536,0.7022746404012045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,16384,0.1547057761086358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,10240,0.06386488676071167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,51200,0.5521262486775717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,12288,0.04708088768853081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,16384,0.1851422256893582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,12288,0.12059733602735732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,8192,0.049845334556367665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,10240,0.041512889994515315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,10240,0.09905955526563857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,12288,0.14261778195699057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,7168,0.04285066657596164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,65536,0.2206231090757582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,8192,0.030878222650951807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,6144,0.03517955541610718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,8192,0.0806844433148702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,7168,0.027064889669418335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,8192,0.09945155514611138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,7168,0.07090400324927436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,5120,0.028805332051383123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,6144,0.024340444140964087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,6144,0.06115644507937961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,4096,0.023545778459972803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,7168,0.08854222297668457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,5120,0.05156444509824117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,4096,0.01763288842307197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,4096,0.04268977708286709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,5120,0.06521688567267524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,4096,0.054413331879509814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,3584,0.01665866706106398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,3584,0.03844533363978068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,3584,0.0491537782880995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,3072,0.03400355577468872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,10240,0.1209004455142551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,2560,0.016880000631014507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,3072,0.015270221564504834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,3072,0.04344177908367581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,2048,0.013819555441538492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,2560,0.014112000664075216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,2560,0.029209779368506536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,5120,0.02056533263789283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,1536,0.011043555206722684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,2560,0.0373982224199507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,2048,0.012770666844315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,2048,0.02479644450876448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,3584,0.021017778250906203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,1024,0.008946666287051307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,1536,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,1536,0.020371556282043457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,768,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,1024,0.015545777148670621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,1024,0.01072355525361167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,1536,0.026271111435360376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,512,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,768,0.013154666456911298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,1024,0.020792000823550753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,512,0.011086222198274402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,512,0.009656888743241629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,256,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,768,0.018569777409235638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,256,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,512,0.016168889072206285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,128,0.0038693332009845306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,256,0.009328888522254096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,6144,0.07563289006551106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,64,0.003504000190231535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,256,0.013796444568369122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,128,0.009323555562231276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,32,0.0035964445107513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,128,0.008616888688670265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,2048,0.03198488884501987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,128,0.011022222538789114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,64,0.008310221963458592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,3072,0.01869777838389079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,32,0.00830577810605367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,65536,0.3137848907046848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,768,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,65536,0.21601688861846924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,16384,0.07714133130179511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,51200,0.17349867026011148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,65536,0.5252239969041612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,51200,0.41342488924662274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,12288,0.060080885887145996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,16384,0.05871466795603434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,65536,0.6954702271355523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,51200,0.548067569732666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,16384,0.18676533963945177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,10240,0.048657778236601085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,12288,0.10242311159769694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,12288,0.04623200164900886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,8192,0.03732266690995958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,12288,0.14152533478207058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,10240,0.040884445110956825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,10240,0.08656266662809585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,7168,0.031601776679356895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,51200,0.23204355769687227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,8192,0.030033777157465618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,8192,0.06985688871807523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,10240,0.1206008858150906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,6144,0.027028444740507338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,7168,0.02637600070900387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,7168,0.06160889069239298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,5120,0.02281422250800663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,8192,0.09853955772187974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,6144,0.02330222229162852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,6144,0.05332444442643059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,16384,0.13459732797410753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,7168,0.0865031083424886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,4096,0.01830666595035129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,5120,0.020405333903100755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,6144,0.07506310939788818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,3584,0.016562667157914903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,5120,0.045443554719289146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,4096,0.017853332890404593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,3072,0.01459822224246131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,4096,0.03728888763321771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,3584,0.016548444827397663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,3584,0.03340533375740051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,3072,0.015202666322390238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,2560,0.01274400038851632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,4096,0.05306933323542277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,3072,0.029618667231665716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,2048,0.010618666807810465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,2560,0.014144889182514615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,3072,0.04182489050759209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,2560,0.02591288917594486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,1536,0.008914666871229807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,2048,0.02178222272131178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,2048,0.012728888955381183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,2560,0.03638133406639099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,1024,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,1536,0.011725333001878528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,1536,0.01806311143769158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,768,0.006071110980378256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,2048,0.031026668018764917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,1024,0.013966222604115805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,1024,0.010741333166758219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,512,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,768,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,768,0.011805333197116852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,1024,0.02036711076895396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,256,0.0038897775941424896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,512,0.010255999863147736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,512,0.009661333428488838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,768,0.01828177769978841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,128,0.003543111185232798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,3584,0.0485155549314287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,256,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,512,0.015503999259736804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,64,0.0032044444233179092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,128,0.009333333207501305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,128,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,32,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,256,0.013431111143694984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,64,0.007639110916190677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,128,0.0107360002067354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,65536,0.22764799329969618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,32,0.007631111476156447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,1536,0.025711999999152288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,51200,0.17857244279649523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,65536,0.1679911083645291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,16384,0.05522488885455661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,5120,0.06439822249942355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,51200,0.14635199970669216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,51200,0.3426524533165826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,12288,0.04016977879736159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,256,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,16384,0.11460355917612712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,65536,0.5541635619269477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,10240,0.035445332527160645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,12288,0.03746577766206529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,51200,0.4345795578426785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,12288,0.08702488740285237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,8192,0.02696977721320258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,10240,0.033538666036393903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,12288,0.11385956075456406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,10240,0.07331999805238512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,7168,0.023451555106374953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,8192,0.05892711215549045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,8192,0.023577777875794306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,10240,0.09827733039855957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,65536,0.4393608835008409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,7168,0.052321778403388135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,7168,0.020632889535692003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,8192,0.08032266961203681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,6144,0.01829777823554145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,16384,0.048839999569786914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,5120,0.01571111049917009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,6144,0.045568890041775174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,5120,0.01608444419172075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,6144,0.06162933508555094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,16384,0.1490808857811822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,4096,0.013092444174819522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,5120,0.03843733337190416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,3584,0.011928000383906893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,4096,0.014161777165200977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,5120,0.05235200126965841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,4096,0.031786666976081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,3072,0.010395555860466426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,3584,0.01310755560795466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,3584,0.02861422300338745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,4096,0.043997334109412305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,2560,0.00906755526860555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,3072,0.01238844460911221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,3072,0.025405334101782903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,6144,0.01888533267709944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,2048,0.007740444607204861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,2560,0.011761777930789523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,3072,0.0340577761332194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,1536,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,7168,0.07014577918582492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,2048,0.018810666269726224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,2048,0.011135111252466837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,1024,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,2560,0.030357334348890517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,1536,0.010724444356229572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,2048,0.025960889127519395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,1536,0.015554666519165039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,1024,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,1024,0.012159110771285163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,1536,0.02167111138502757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,512,0.003834666477309333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,768,0.01054222219520145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,768,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,1024,0.017509332961506315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,256,0.003289777785539627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,512,0.00868977771864997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,512,0.008979555633332994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,3584,0.03858133488231235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,128,0.0029004443850782183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,256,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,256,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,512,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,2560,0.022128888302379187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,64,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,32,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,128,0.00830488900343577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,256,0.012103999654452005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,128,0.007157333195209503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,65536,0.14736799399058023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,64,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,32,0.006593777901596493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,768,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,51200,0.119832886589898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,65536,0.16546577877468532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,16384,0.03575111097759671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,65536,0.35511289702521426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,51200,0.13282844755384657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,51200,0.27950043148464626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,12288,0.027224888404210407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,768,0.01585244470172458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,16384,0.04669422242376539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,16384,0.09440888961156209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,10240,0.02218577762444814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,12288,0.0376506679587894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,51200,0.4200115468766954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,16384,0.14465422100490993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,12288,0.07116711139678955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,8192,0.016347555650605094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,10240,0.033077332708570696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,10240,0.059808002577887646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,12288,0.11155200004577637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,7168,0.013470222552617392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,128,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,8192,0.022583999567561682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,8192,0.048487110270394214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,6144,0.011565333439244164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,10240,0.09528533617655437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,7168,0.02028622229894002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,7168,0.04295555419392056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,5120,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,6144,0.018509333332379658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,6144,0.0376800000667572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,7168,0.06714489062627156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,4096,0.008365333080291748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,5120,0.016413332687483895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,5120,0.03220977716975742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,6144,0.0591848889986674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,3584,0.007784888976150089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,4096,0.014440000057220459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,5120,0.05062133404943678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,4096,0.026547556122144062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,3072,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,4096,0.04145244426197476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,3584,0.023577777875794306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,3584,0.013186666700575085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,2560,0.006206222292449739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,3072,0.012448888685968188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,3072,0.020988444487253826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,3584,0.03734044565094842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,2048,0.005325333525737126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,2560,0.01849688920709822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,3072,0.0332568883895874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,8192,0.07726666662428114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,1536,0.0046311111913787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,2048,0.01587288909488254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,2048,0.011400000088744693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,2560,0.02922399838765462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,1024,0.0038746665749284956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,1536,0.010372444159454769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,2048,0.025126222107145522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,1536,0.013185777597957186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,768,0.0034942221310403612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,65536,0.5341608789232042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,1024,0.010747555229398938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,1024,0.009665778113736046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,1536,0.021055110626750525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,512,0.002991111121243901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,768,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,768,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,256,0.0025928889711697898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,1024,0.01720888912677765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,512,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,512,0.008734222087595198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,128,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,768,0.01551822159025404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,256,0.008349333372380998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,256,0.006315555423498154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,2560,0.012039110892348819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,64,0.0022515555222829184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,32,0.0024666666156715816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,128,0.008349333372380998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,128,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,256,0.011545778148704104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,65536,0.09780177805158828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,128,0.009008888569143083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,64,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,32,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,51200,0.07824266619152494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,16384,0.02951644526587592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,65536,0.16597421964009604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,65536,0.31538489129808217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,51200,0.13323377238379583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,12288,0.022239999638663396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,51200,0.24745599428812662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,16384,0.08408355712890625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,16384,0.047240889734692044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,10240,0.015980444020695157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,12288,0.03933777742915683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,65536,0.5276044209798177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,51200,0.41119734446207684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,16384,0.14260266886817083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,8192,0.01312888910373052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,512,0.013478221992651621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,10240,0.05315733287069532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,12288,0.11029777924219768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,10240,0.03414400087462531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,7168,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,8192,0.024347555306222703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,6144,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,8192,0.043324444029066295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,7168,0.020959110723601446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,10240,0.09412622451782227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,7168,0.03803199860784743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,5120,0.008707555631796518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,8192,0.076255997021993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,6144,0.018838221828142803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,4096,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,7168,0.06705511278576322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,6144,0.03338222371207343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,5120,0.01681777834892273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,3584,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,5120,0.028648889727062646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,6144,0.057982219590081104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,4096,0.014500444134076437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,3072,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,4096,0.02369244396686554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,5120,0.049754665957556836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,3584,0.013827555709415011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,2560,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,3584,0.02139733400609758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,4096,0.0407315558857388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,12288,0.06296000215742323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,3072,0.019270223047998216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,3584,0.03657955593532986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,2560,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,2560,0.016903999778959487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,1536,0.004171555654870139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,2048,0.011424889167149862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,2048,0.014720888601409065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,2560,0.0284862220287323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,2048,0.02442311081621382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,1536,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,1536,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,1536,0.020654223031467862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,1024,0.009913778139485253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,768,0.0034560000316964257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,1024,0.009674666656388177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,1024,0.01647555496957567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,3072,0.01311377767059538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,768,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,768,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,2048,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,512,0.00869955536392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,3072,0.03288088904486762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,768,0.01510933372709486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,512,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,256,0.008424888882372115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,512,0.012452444268597497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,128,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,256,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,1024,0.003526222374704149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,64,0.0021342221233579847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,128,0.008316444853941599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,32,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,128,0.00544177782204416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,65536,0.08858488665686713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,64,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,51200,0.07306222120920818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,32,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,16384,0.02808533443344964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,512,0.002809777855873108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,12288,0.021567111214001972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,10240,0.01703288820054796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,65536,0.31719912423027885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,256,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,51200,0.24727111392550996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,16384,0.08313244581222534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,8192,0.012684444586435953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,7168,0.011060444845093621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,6144,0.009848000274764167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,12288,0.0629040002822876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,5120,0.00831911133395301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,4096,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,10240,0.052975111537509494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,3584,0.006568888823191325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,8192,0.042938666211234204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,7168,0.037744889656702675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,3072,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,6144,0.03296000096533034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,5120,0.028121779362360638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,2560,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,256,0.01107377807299296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,2048,0.00452622233165635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,1536,0.003919110943873723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,4096,0.023709333605236475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,1024,0.0032231110251612137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,768,0.003044444446762403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,3584,0.02129688858985901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,512,0.0025031111306614345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,3072,0.01896711190541585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,2560,0.016699555847379897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,256,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,2048,0.014493332968817817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,128,0.002116444417172008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,1536,0.012171555724408893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,64,0.001968888979819086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,1024,0.009741333623727163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,32,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,768,0.008384000096056197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,512,0.007510222494602203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,65536,0.09495911333296035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,256,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,128,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,64,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,16384,0.0266951107316547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,32,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,12288,0.020145777199003432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,10240,0.01200799975130293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,8192,0.010025777750545079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,65536,0.31647910012139213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,16384,0.08296533425649007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,12288,0.06273511383268568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,51200,0.24702665540907118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,7168,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,6144,0.008343111309740279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,5120,0.007840000092983246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,4096,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,8192,0.043246222866906055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,3584,0.006675555474228329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,7168,0.03813599877887302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,6144,0.033003555403815374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,5120,0.028188443846172754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,3072,0.011720000041855706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,2560,0.010194666683673859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,4096,0.023616888456874426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,2048,0.004518222063779831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,1536,0.00387999994887246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,3584,0.021988444858127173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,3072,0.019740444090631273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,2560,0.016869333055284288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,1024,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,768,0.002806222273243798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,1536,0.012364444633324942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,51200,0.07476266887452868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,512,0.002499555548032125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,256,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,1024,0.009886222581068674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,768,0.008355555435021719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,128,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,64,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,512,0.007548444800906711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,256,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,10240,0.05277688966857063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,128,0.005294222384691238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,64,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,32,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,2048,0.014832889040311178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,16384,1.7958026462131078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,12288,4.283471001519097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,16384,5.230157216389974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,16384,5.892926957872178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,12288,1.3028862211439345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,12288,4.3267974853515625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,51200,7.0934354994032125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,10240,1.1426915062798393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,10240,3.5804727342393665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,8192,0.9050363964504666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,10240,3.425785700480143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,8192,2.818360858493381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,7168,0.7722782029045953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,8192,2.7600782182481556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,7168,2.4048444959852433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,6144,0.6498026847839355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,7168,2.3692614237467446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,6144,2.0277351803249783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,5120,0.5505706469217936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,6144,2.068666670057509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,5120,1.6510018242730036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,4096,0.4587430953979492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,5120,1.7096391254001195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,4096,1.4665999942355687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,3584,1.1542861726548936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,3584,0.41784000396728516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,4096,1.3499582078721788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,3072,0.3599991003672282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,3072,1.0737359788682725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,3584,1.169774267408583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,2560,0.8719075520833334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,3072,0.9733724594116211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,2560,0.32258399327596027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,2048,0.65810579723782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,2048,0.26925423410203725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,2560,0.8747982449001737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,1536,0.5557040108574761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,2048,0.7448888884650336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,1536,0.21415555477142334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,1024,0.3325822353363037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,1536,0.5730444590250651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,1024,0.1677466630935669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,768,0.28686311509874135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,512,0.19641955693562826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,51200,17.14947001139323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,768,0.14717333846622044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,1024,0.49312533272637266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,256,0.1583626667658488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,768,0.34946666823493105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,512,0.1395217842525906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,256,0.11924089325798883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,128,0.141503996319241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,512,0.35108266936408145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,256,0.24032976892259386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,128,0.11590666241115993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,64,0.10138399733437432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,32,0.1221697727839152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,128,0.23127910825941297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,128,0.008282666405042013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,32,0.23783734109666613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,64,0.23820355203416613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,16384,4.031002468532987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,51200,5.390067630343967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,51200,25.625773111979168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,65536,7.203719245062934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,16384,1.4200870725843642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,16384,4.5684996710883246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,12288,1.028341293334961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,12288,3.113432990180122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,10240,2.8357662624782987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,10240,0.8163502481248645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,12288,3.5932282341851125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,51200,12.950590345594618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,8192,0.6585040092468262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,8192,2.2050613827175565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,10240,2.9264034695095487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,8192,2.1618124643961587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,7168,2.1795982784695096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,7168,1.7884381612141926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,7168,0.601753764682346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,6144,0.5557529131571451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,6144,1.8421226077609594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,6144,1.5453404320610895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,65536,16.73855251736111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,5120,1.3860639995998805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,5120,0.4958702193366156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,4096,0.36118398772345645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,4096,1.0331510967678494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,5120,1.3327350616455078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,3584,0.3287031120724148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,3584,0.8914320203993055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,4096,0.9942471186319987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,3072,0.7744435734219022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,3072,0.2887929015689426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,3584,0.9214284684922961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,3072,0.7803209092881945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,2560,0.2586133215162489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,2560,0.6861964331732856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,2048,0.21044622527228463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,2560,0.6472791035970052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,2048,0.5679164462619358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,1536,0.1711039940516154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,1536,0.43205867873297793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,1024,0.27166223526000977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,768,0.2399680084652371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,1024,0.13448177443610296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,1536,0.4594728681776259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,512,0.17618578010135225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,1024,0.5012044376797146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,768,0.15442132949829102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,256,0.1603777805964152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,512,0.11355466312832302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,256,0.09205600288179185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,256,0.25835466384887695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,128,0.08983466360304092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,2048,0.533532460530599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,64,0.08034666379292806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,128,0.09346044725841945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,32,0.08481778038872613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,128,0.1789262162314521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,51200,20.85664537217882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,64,0.18326400385962593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,32,0.18755822711520723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,768,0.2662008868323432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,512,0.210554665989346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,16384,1.2848684522840712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,51200,1.6756160524156358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,65536,2.3657635582817926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,16384,0.44038844108581543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,16384,1.4963066313001845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,12288,0.3308151033189562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,12288,1.2943706512451172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,12288,1.088370641072591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,51200,4.622547573513455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,10240,0.2780800130632189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,10240,0.8623218006557889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,65536,25.769336276584202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,8192,0.25343200895521373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,10240,0.9343066745334201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,65536,5.670696682400173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,8192,0.6490800115797255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,7168,0.6501902474297417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,8192,0.6933697594536675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,7168,0.2007573313183255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,6144,0.5455644395616319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,6144,0.22511021296183267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,7168,0.6361795531378852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,5120,0.4451671176486545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,6144,0.5608622233072916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,5120,0.42345333099365234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,5120,0.18023110760582817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,4096,0.35734576649136013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,4096,0.13209954897562662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,3584,0.2812799877590603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,51200,6.613754696316189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,4096,0.3359644412994385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,3584,0.13696800337897405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,3072,0.2710008886125353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,3072,0.10372177759806316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,3584,0.2927440007527669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,2048,0.2006631162431505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,2560,0.22243555386861166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,3072,0.2685848871866862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,2560,0.10993955532709758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,2048,0.07247733407550387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,2560,0.21708444754282633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,1536,0.15179733435312906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,1536,0.05908888578414917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,768,0.0866924458079868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,2048,0.18492089377509224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,1024,0.04854933420817057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,1536,0.1403368843926324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,512,0.0520977775255839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,256,0.035735110441843666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,768,0.05561688873502943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,512,0.0420595539940728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,768,0.08578755458196004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,1024,0.1065431104765998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,512,0.07231466637717353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,64,0.026318222284317017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,256,0.06576622194714017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,128,0.03635911146799723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,32,0.028135110934575398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,128,0.06126844220691257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,65536,8.499802483452692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,64,0.06285778019163343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,32,0.06497244702445136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,1024,0.09185866514841716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,128,0.03162311183081733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,65536,1.6399795744154189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,256,0.04713066750102573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,51200,1.1768506368001301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,16384,0.4367893271976047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,16384,1.2324977450900607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,51200,3.191813362969292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,12288,0.696772469414605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,16384,1.138594627380371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,12288,0.2927324506971571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,10240,0.21696889400482178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,10240,0.6847110854254829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,12288,0.8388551076253256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,65536,4.329394870334201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,10240,0.6945084465874566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,8192,0.6261848873562282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,8192,0.5124044418334961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,8192,0.2222773234049479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,7168,0.42030488120185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,7168,0.15311377578311497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,6144,0.13518667221069336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,5120,0.3267884519365099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,6144,0.5152773327297634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,7168,0.5080302026536729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,6144,0.38303467962476945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,5120,0.11384533511267768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,5120,0.3376364443037245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,4096,0.2674560017055935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,4096,0.09504533476299709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,3584,0.23282933235168457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,3584,0.08467111322614883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,3072,0.2002506653467814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,3584,0.22809155782063803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,3072,0.07710222403208415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,2560,0.17663110627068412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,3072,0.1943662166595459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,2048,0.13839111063215467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,2560,0.06650311417049833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,1536,0.09922311041090225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,2560,0.19161778026156953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,2048,0.0565662251578437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,1024,0.07076799869537354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,2048,0.14349155955844456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,65536,6.327871110704209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,1536,0.0510808891720242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,768,0.060347557067871094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,1536,0.10883733299043442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,512,0.04159377680884467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,768,0.03958666655752394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,512,0.03376355436113145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,256,0.0303422212600708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,512,0.05623200204637316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,768,0.07088088989257812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,256,0.030393779277801514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,128,0.023807111713621352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,64,0.020421332783169217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,256,0.05128711130883959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,128,0.03000533249643114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,51200,5.569445292154948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,32,0.021500445074505273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,128,0.04899377624193827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,64,0.0487973325782352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,32,0.050591111183166504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,1024,0.04116266634729173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,1024,0.08242044183943006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,4096,0.42585065629747176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,16384,0.826506667666965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,65536,1.3446986940171983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,51200,1.0412071016099718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,16384,0.29720266660054523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,16384,0.9552364349365234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,12288,0.22214310699039033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,12288,0.7657884491814507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,51200,2.7027511596679688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,12288,0.7527582380506727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,10240,0.580293337504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,10240,0.5665475527445475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,8192,0.4277697669135199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,65536,3.4032995435926647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,8192,0.4392746554480658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,7168,0.13131999969482422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,7168,0.39925511678059894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,6144,0.3147137694888645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,7168,0.4596675766838922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,6144,0.14685332775115967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,5120,0.2613404326968723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,6144,0.3336026668548584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,5120,0.10261333651012844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,5120,0.27498043908013237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,4096,0.20413245095147026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,4096,0.08093866374757555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,8192,0.20952178372277153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,3584,0.195978668000963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,4096,0.21778221925099692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,3584,0.07286577754550509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,3072,0.16112089157104492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,3584,0.19743200143178305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,3072,0.06465866830613878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,2560,0.13268978065914577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,65536,5.232831319173177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,3072,0.1651582188076443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,2048,0.11903022395239936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,1536,0.08859199947781032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,2560,0.14211022853851318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,51200,4.032566918267144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,2048,0.1463413370980157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,1536,0.040912889771991305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,1536,0.10550044642554389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,1024,0.03480799992879232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,768,0.04845244354671902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,1024,0.07386666536331177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,512,0.03936266567971971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,768,0.032446222172843084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,768,0.056736886501312256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,512,0.030220445659425523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,512,0.048566222190856934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,2560,0.056899554199642606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,256,0.02671555512481266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,128,0.018223999275101554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,256,0.04303555687268575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,2048,0.04896977874967787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,64,0.017542221479945712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,32,0.018565333551830716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,10240,0.21141421794891357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,1024,0.06090844339794583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,128,0.04069777660899692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,64,0.041462222735087075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,32,0.042809777789645724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,256,0.02514666650030348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,128,0.0262444449795617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,16384,0.5867057906256782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,65536,1.2759084701538086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,16384,0.25726132922702366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,51200,0.9447884029812283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,16384,0.743560896979438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,12288,0.23638399442036948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,51200,2.354212443033854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,12288,0.8692293167114258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,65536,2.718977822197808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,10240,0.4268888897365994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,12288,0.5890773137410482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,10240,0.17606843842400444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,8192,0.38823554250929093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,7168,0.2775431209140354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,10240,0.6404862403869629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,7168,0.1302408907148573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,6144,0.24244888623555502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,8192,0.3681137826707628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,7168,0.3307733270857069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,6144,0.12063644991980658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,5120,0.08409866359498765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,51200,3.530580520629883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,6144,0.26947111553615993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,4096,0.19516800509558785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,5120,0.22132533126407197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,3584,0.1591457790798611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,4096,0.07479200098249647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,3584,0.06268977456622653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,4096,0.17949867248535156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,3072,0.1162231100930108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,65536,4.216250525580512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,2560,0.10609066486358643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,3584,0.16510132948557535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,8192,0.13908444510565862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,3072,0.05560355716281467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,3072,0.1414239936404758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,1536,0.06848800182342529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,2560,0.05340977840953403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,2048,0.04154310954941644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,1024,0.04630044433805677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,1536,0.034276445706685386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,768,0.03870133227772183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,1536,0.07835911379920112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,1024,0.030744890371958416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,512,0.02737599942419264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,5120,0.20748444398244223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,1024,0.0579893324110243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,768,0.0262746661901474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,256,0.01906577746073405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,768,0.046911110480626426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,512,0.038839999172422625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,256,0.02109333376089732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,128,0.016580444243219163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,64,0.014089778065681458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,256,0.03437066740459866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,32,0.015093333191341825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,128,0.021176889538764954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,128,0.03250933355755276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,64,0.03196177879969279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,2048,0.09341688950856526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,32,0.032058666149775185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,2560,0.11766666836208767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,2048,0.09638310803307427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,512,0.02362577782736884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,16384,0.5040168762207031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,65536,1.038247956169976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,51200,0.8114168908860949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,16384,0.3108888996971978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,12288,0.5364301999409994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,51200,2.1050569746229386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,16384,0.6771395471360949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,12288,0.24543110529581705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,10240,0.4505804379781087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,65536,2.7130461798773875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,10240,0.13873244656456843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,12288,0.47756269243028426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,8192,0.3463111188676622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,10240,0.42320889896816677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,7168,0.2607653405931261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,7168,0.0992106662856208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,51200,2.9991590711805554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,6144,0.22312977578904894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,8192,0.4239635467529297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,7168,0.2746693293253581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,5120,0.19688267178005645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,6144,0.08772444725036621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,4096,0.15944622622595891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,5120,0.0921635561519199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,3584,0.1178800000084771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,4096,0.06253600120544434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,5120,0.2001111110051473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,65536,3.6660711500379772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,3072,0.11682132879892985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,4096,0.1703022188610501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,3584,0.05601244502597385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,3584,0.13926666312747532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,3072,0.05161688725153605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,2560,0.10331644614537557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,2048,0.07811911238564385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,3072,0.12235377894507514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,2560,0.044053332673178784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,1536,0.058840003278520375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,1024,0.045477333996031016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,8192,0.1344168848461575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,2560,0.11114845010969375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,768,0.03348888953526815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,2048,0.09082221984863281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,1024,0.02567022211021847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,1536,0.06827021969689263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,512,0.023769777682092454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,768,0.02453688946035173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,256,0.018065777089860704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,768,0.042564445071750216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,512,0.020969778299331665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,6144,0.24092443784077963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,128,0.013947554760509066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,512,0.03394844465785556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,256,0.01936533384852939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,64,0.012030222349696688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,256,0.0299271113342709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,32,0.012401777837011548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,128,0.0188426673412323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,128,0.028480887413024902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,64,0.028265777561399672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,32,0.028149333265092637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,2048,0.03699288765589396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,1536,0.031123555368847314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,1024,0.05156977640257942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,16384,0.43724621666802305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,65536,0.8680062294006348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,51200,0.6812728775872124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,16384,0.19125066863165963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,16384,0.5690293312072754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,12288,0.1427626609802246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,12288,0.5080622037251791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,51200,1.698129759894477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,10240,0.34186045328776044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,65536,2.226031197441949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,12288,0.4759039878845215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,8192,0.28886577818128795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,10240,0.35229333241780597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,7168,0.21480354997846815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,8192,0.11883466773562962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,8192,0.375852452384101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,6144,0.17283288637797037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,7168,0.08765688869688247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,7168,0.3053377734290229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,51200,2.5110905965169272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,5120,0.16715289486779106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,6144,0.08376977841059367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,6144,0.235551118850708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,5120,0.06643288665347628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,4096,0.05640800131691826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,3584,0.12204533153110081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,5120,0.1731466717190213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,3072,0.09648178021113078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,4096,0.14486043983035618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,3584,0.0506533318095737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,65536,3.399223963419596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,3584,0.12267643875545925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,2560,0.08887200223075019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,10240,0.15386933750576443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,2048,0.06841066810819837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,3072,0.10628533363342285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,3072,0.04497244291835361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,2560,0.04481066597832573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,1536,0.05312711000442505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,1024,0.03772888912094964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,2048,0.033787555164761014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,2048,0.07618755764431424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,768,0.02805866797765096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,2560,0.0919182234340244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,1024,0.023398222194777593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,512,0.02328355610370636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,1024,0.0456026660071479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,1536,0.060221334298451744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,768,0.021223111285103693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,768,0.03751111030578613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,512,0.019182221757041085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,256,0.016131555040677387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,4096,0.13880977365705702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,128,0.012823999755912356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,512,0.03018933203485277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,64,0.011969777444998423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,256,0.017423111531469557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,128,0.017249777913093567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,64,0.024995555480321247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,32,0.024752888414594863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,128,0.024911110599835713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,1536,0.03250755535231696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,65536,0.7446373303731283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,32,0.012724444270133972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,65536,1.6268284055921767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,256,0.02625511089960734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,16384,0.3599955505794949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,51200,1.5023537741767035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,16384,0.16681599617004395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,12288,0.2943386766645643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,51200,0.6154986487494575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,10240,0.24682845009697807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,12288,0.12662933932410345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,16384,0.510517332288954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,10240,0.11581333478291829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,8192,0.19814133644104004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,10240,0.3228000005086263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,12288,0.41961510976155597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,8192,0.08437599738438924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,7168,0.200582226117452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,51200,2.0703023274739585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,7168,0.07646577888064914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,6144,0.17748978402879503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,5120,0.13360622194078234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,65536,2.781713697645399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,6144,0.06954044765896268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,4096,0.1008346676826477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,7168,0.21831022368537056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,6144,0.1771991120444404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,5120,0.05863200293646919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,3584,0.08823288811577691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,4096,0.0502053333653344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,3072,0.07960266537136503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,5120,0.14966577953762478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,3584,0.043403555949529014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,4096,0.12670044104258218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,2560,0.068722665309906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,3072,0.03955377803908454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,2048,0.0571991138988071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,3584,0.10658844312032063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,1536,0.041285332706239485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,2560,0.033887111478381686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,3072,0.11133778095245361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,2048,0.02971200148264567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,2560,0.07953066958321466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,1536,0.02446844511561924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,768,0.022145777940750122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,2048,0.06634222136603461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,1536,0.05197244551446703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,512,0.016679111454221938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,1024,0.02055644492308299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,256,0.012328888807031842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,1024,0.039255112409591675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,768,0.018792000081804063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,128,0.010301333334710863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,512,0.017112889223628573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,768,0.03270666797955831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,64,0.009073778159088558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,256,0.01577422188387977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,512,0.026000888811217412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,32,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,128,0.016007999579111736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,128,0.02163822286658817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,8192,0.23017422358194986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,64,0.021643555826610986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,32,0.02155911094612545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,1024,0.030304888884226482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,65536,0.5889004601372613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,51200,0.45882309807671445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,16384,0.2966586748758952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,51200,1.0994266933865018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,16384,0.13885422547658285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,256,0.02237244447072347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,65536,1.5057377285427516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,12288,0.11051199833552043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,16384,0.45844623777601456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,10240,0.09018133083979289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,12288,0.3496675491333008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,10240,0.24404888682895234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,8192,0.16209867265489367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,65536,2.1879475911458335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,8192,0.07113422287835015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,7168,0.1353706651263767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,6144,0.1174960003958808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,7168,0.06372000111473931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,8192,0.1928782198164198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,12288,0.2271955543094211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,6144,0.056461334228515625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,10240,0.21821067068311903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,6144,0.14754045009613037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,7168,0.18014132976531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,4096,0.08584533135096233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,5120,0.12327822049458821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,3584,0.07925333579381307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,4096,0.04032622112168206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,3072,0.06748355759514703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,51200,1.6946159998575847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,4096,0.10050755739212036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,3584,0.03528533379236857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,3584,0.0886426634258694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,2560,0.05927822324964735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,3072,0.03238933285077413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,2048,0.04216266671816508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,2560,0.02864977717399597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,3072,0.07833955685297649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,1536,0.03314310974544949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,2048,0.025052444802390203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,2560,0.06569244464238484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,1024,0.022265776991844177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,2048,0.05526489019393921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,1536,0.02127644419670105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,768,0.01886755559179518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,1024,0.018252443936136033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,1536,0.0435697767469618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,1024,0.03312622176276313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,5120,0.10993778043323094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,768,0.01643288963370853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,256,0.010300444232092964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,128,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,512,0.015000000596046448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,768,0.026943999860021803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,256,0.014117333624098035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,512,0.021869333253966436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,64,0.007677333222495184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,32,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,128,0.014379555980364481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,256,0.018917333748605516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,128,0.01828977796766493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,64,0.017875555488798354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,32,0.01809866726398468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,65536,0.5292586750454372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,51200,0.416777769724528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,51200,0.8142302301194934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,65536,1.0543609195285373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,16384,0.31320712301466197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,5120,0.048841777775022716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,16384,0.1328586604860094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,12288,0.2244755427042643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,12288,0.09674933221605088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,16384,0.42464176813761395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,12288,0.30757066938612193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,8192,0.15554400285085043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,10240,0.08199555344051786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,51200,1.3175172805786133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,10240,0.24008443620469835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,7168,0.14134844144185385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,65536,1.8951216803656683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,8192,0.06963911321428087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,6144,0.0959191123644511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,7168,0.05967555443445841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,7168,0.1580586698320177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,8192,0.17605333858066133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,5120,0.08378755384021336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,6144,0.0536080002784729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,4096,0.07902222209506564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,6144,0.14699110719892713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,5120,0.04396533303790622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,3584,0.06483199861314562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,4096,0.03673155440224542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,3072,0.056977775361802846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,4096,0.09063377645280625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,3584,0.032666666640175715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,2560,0.04623288909594218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,3072,0.028963555892308552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,3584,0.08143733607398139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,2048,0.042005333635542125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,3072,0.0689386659198337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,10240,0.19222754902309844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,1536,0.029532445801628962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,2560,0.02637600070900387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,1024,0.019298666053348117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,2560,0.05934400028652615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,2048,0.022463111413849726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,768,0.016385777129067313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,1536,0.01904177831278907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,2048,0.04977333214547899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,1024,0.016321778297424316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,512,0.013174222575293647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,1024,0.029747555653254192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,768,0.015558222929636637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,256,0.009313777916961247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,768,0.02479022244612376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,512,0.014131555954615274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,512,0.014373333917723762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,512,0.019819556011093985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,128,0.008006222546100616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,64,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,256,0.013415999710559845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,256,0.01737955543729994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,32,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,128,0.013165333204799227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,5120,0.1139448881149292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,128,0.01682133310370975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,64,0.016525333126386006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,32,0.016368889146380954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,65536,0.45796887079874676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,51200,0.36368799209594727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,51200,0.681754642062717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,16384,0.26693688498602974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,65536,1.0951627095540364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,16384,0.10788710912068684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,12288,0.21528800328572592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,16384,0.33815821011861164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,12288,0.084279113345676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,51200,1.1816160413953993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,10240,0.1682222286860148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,12288,0.3115688959757487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,10240,0.07196888658735487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,8192,0.14099021752675375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,10240,0.1950328879886203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,7168,0.13047377268473306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,8192,0.05999466445710924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,1536,0.039572444227006696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,6144,0.09194844298892552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,7168,0.0517484446366628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,7168,0.1518053346210056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,5120,0.08428088823954265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,6144,0.04518666532304552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,4096,0.06270488765504625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,6144,0.11782756116655137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,5120,0.03973955578274197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,3584,0.05726310941908094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,5120,0.09950755702124701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,4096,0.03187377916442023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,3584,0.029623111089070637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,2560,0.04064177804523044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,3584,0.070814225408766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,3072,0.06214844518237644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,2560,0.02374222212367588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,2048,0.03604622350798713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,65536,1.583989355299208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,2560,0.053583999474843345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,1536,0.02738044493728214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,2048,0.020655110478401184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,1024,0.018757333358128864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,1536,0.018106665876176622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,8192,0.16464000278049046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,1536,0.03597955571280585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,768,0.015135111080275642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,1024,0.015489778584904142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,512,0.01168444421556261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,1024,0.027033777700530157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,768,0.022527111901177302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,512,0.013266666895813413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,256,0.008560000194443597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,3072,0.04667466547754076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,512,0.018022222651375663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,4096,0.07960800329844157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,256,0.012431110772821637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,3072,0.02630666726165348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,64,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,32,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,128,0.01257955531279246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,128,0.01514933341079288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,64,0.01481066644191742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,32,0.014837332897716098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,2048,0.0446648895740509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,768,0.014456889695591398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,65536,0.4071546660529242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,65536,0.942928844028049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,51200,0.6612222459581163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,128,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,16384,0.24362132284376356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,51200,0.3273591200510661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,256,0.01574222246805827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,12288,0.14852978123558894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,65536,1.1638853285047743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,10240,0.14611111746893987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,12288,0.08426844411426121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,51200,0.9027119742499458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,16384,0.3586533334520128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,8192,0.12242666880289714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,12288,0.2144115500979953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,10240,0.06708621978759766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,7168,0.10167644421259563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,10240,0.16671555572085908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,8192,0.05264622304174635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,6144,0.08714933527840509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,7168,0.046775112549463906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,8192,0.13440444734361437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,5120,0.07004800107744005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,6144,0.04011733333269755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,7168,0.12081066767374675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,4096,0.05188088946872287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,3584,0.04498844345410665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,5120,0.0343466666009691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,6144,0.10165155596203274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,5120,0.08705688847435845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,3584,0.02608000073168013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,4096,0.07069866524802314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,2560,0.0355057782597012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,3584,0.06305511130226983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,3072,0.023617777559492324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,16384,0.11861156092749701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,2048,0.02844711144765218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,2560,0.02128177715672387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,1536,0.020610666937298246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,2560,0.04642311069700453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,2048,0.018495111001862418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,1024,0.01498222185505761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,2048,0.03944177760018243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,1536,0.016068443655967712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,768,0.012039999994966718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,1536,0.03146044413248698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,1024,0.014442667365074158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,512,0.009631111390060848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,1024,0.023931556277804907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,768,0.013577777478430005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,768,0.019824888971116807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,256,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,512,0.012416889270146688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,128,0.006167111297448476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,512,0.016528000434239704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,4096,0.028532445430755615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,64,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,256,0.01198755535814497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,32,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,128,0.011710222396585675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,256,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,128,0.013445333474212222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,64,0.013066666821638743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,3072,0.053852445549435086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,32,0.013112000293201871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,16384,0.17193155818515352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,65536,0.32596977551778156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,51200,0.46278932359483504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,65536,0.6474266582065159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,12288,0.13342755370669895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,16384,0.08218844731648763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,10240,0.10955911212497288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,12288,0.06747911373774211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,12288,0.20830133226182726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,51200,0.7578248977661133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,8192,0.08719199895858765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,10240,0.14559733867645264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,8192,0.04532533221774631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,7168,0.07858666446473864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,8192,0.11742311053805882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,6144,0.07040533092286852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,7168,0.03904977771970961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,7168,0.10270044538709854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,6144,0.03397422366672092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,5120,0.05580177572038439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,6144,0.08838399913575913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,4096,0.044906665881474815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,5120,0.02993511160214742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,16384,0.2284026675754123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,5120,0.07420000102784899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,3584,0.0413351125187344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,4096,0.025217778152889673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,3072,0.04062666826777988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,3072,0.03448977735307481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,4096,0.06009422408209907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,10240,0.05836533175574409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,3584,0.022839110758569505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,3072,0.020972443951500788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,2048,0.022703111171722412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,2560,0.018961777289708454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,3072,0.04741777645217048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,1536,0.018421333697107103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,2048,0.016916443904240925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,2048,0.034838222795062594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,1024,0.01314755529165268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,1536,0.015146666102939181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,51200,0.26201510429382324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,768,0.01069777790043089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,1024,0.01309866706530253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,1024,0.020940444535679285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,512,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,768,0.012298666768603854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,768,0.017530666457282174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,65536,0.9387315114339193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,512,0.01144977741771274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,512,0.014482667048772177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,2560,0.027097778187857732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,3584,0.05353688862588671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,128,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,256,0.010945777926180096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,64,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,256,0.0123848890264829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,32,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,128,0.011011555790901184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,2560,0.04103733433617486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,128,0.011750222080283694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,64,0.011751111182901593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,32,0.011648000114493899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,1536,0.027642667293548584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,65536,0.5218142403496636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,65536,0.26791734165615505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,51200,0.4020204544067383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,51200,0.21339910560184053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,16384,0.14869244893391928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,256,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,16384,0.07264710797203912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,12288,0.10091555780834621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,65536,0.754110230339898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,10240,0.09480977720684475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,16384,0.19581866264343262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,51200,0.58151822619968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,12288,0.061221334669325084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,8192,0.07354311148325603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,12288,0.16973511377970377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,7168,0.06370577547285292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,8192,0.040056887600157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,10240,0.11873155170016819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,6144,0.05125777920087179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,5120,0.042971554729673594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,8192,0.09695288870069717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,7168,0.037441776858435735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,4096,0.03505511085192362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,6144,0.032016889916525945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,5120,0.027503111296229895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,6144,0.07194844219419691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,3584,0.03033422099219428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,4096,0.023250665929582384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,5120,0.060870223575168185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,3072,0.02552622225549486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,4096,0.050159109963311084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,3584,0.02010844482315911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,2560,0.022728888524903193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,3072,0.018395556343926322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,3584,0.04459111226929558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,2048,0.018408000469207764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,3072,0.03962488969167074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,1536,0.014671110444598727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,2048,0.015807110402319167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,1024,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,1536,0.013452444639470843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,10240,0.05650577942530314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,1536,0.023482667075263128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,768,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,1024,0.011740444435013665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,1024,0.017948443690935772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,512,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,768,0.011391999820868174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,256,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,7168,0.08588088883294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,768,0.015154666370815702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,128,0.004616888860861461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,512,0.012433778080675336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,256,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,64,0.004219555606444676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,128,0.01035199976629681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,32,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,128,0.010203555226325989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,2560,0.017490666773584154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,64,0.00982844415638182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,32,0.009992000129487779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,2560,0.03460089034504361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,65536,0.36817510922749835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,2048,0.028798222541809082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,65536,0.2149946689605713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,16384,0.09637511438793606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,51200,0.27812356419033474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,51200,0.17776889271206328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,12288,0.06605066855748494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,16384,0.057554668850368924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,512,0.010386666489972008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,10240,0.06379910972383288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,65536,0.5917510986328125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,16384,0.1531253390842014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,8192,0.052128887838787503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,12288,0.12130133310953777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,256,0.010695111420419482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,7168,0.04275555411974589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,10240,0.09576800134446885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,8192,0.02985066506597731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,8192,0.07748266723420885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,6144,0.03748888770739237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,7168,0.02750844425625271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,5120,0.02885422110557556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,7168,0.06809155808554755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,6144,0.023608888188997906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,4096,0.022321777211295232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,5120,0.020419556233617995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,6144,0.0588888857099745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,3584,0.019848888119061787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,5120,0.04985688792334663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,4096,0.017657778329319425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,3072,0.017702221870422363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,4096,0.04126222266091241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,3584,0.016473778420024447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,2560,0.015281778242852954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,3584,0.03719910979270935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,3072,0.01516266663869222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,2048,0.012703999876976013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,12288,0.04582666688495212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,3072,0.03232444326082865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,2560,0.013825777504179211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,1536,0.010479110810491772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,10240,0.04026044408480326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,1024,0.008023111356629265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,2048,0.023959111836221483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,1536,0.019650666250122916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,768,0.006753777878151999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,1024,0.010597333312034607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,1024,0.015075556106037565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,512,0.005395555661784277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,768,0.010313777459992303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,768,0.01261333293384976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,256,0.004524444540341695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,512,0.009843555589516958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,512,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,128,0.003834666477309333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,256,0.009110222260157267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,256,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,64,0.00350577798154619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,128,0.009022222624884712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,51200,0.4603448973761664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,32,0.003527111063400904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,128,0.008631111019187504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,2560,0.02825333343611823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,64,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,1536,0.011454222102959951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,32,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,65536,0.28894133037990993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,51200,0.2340559959411621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,16384,0.07737777630488078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,65536,0.20917956034342447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,12288,0.05745600329505073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,51200,0.16516710652245417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,16384,0.06387466854519315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,10240,0.04783644609981113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,65536,0.5042808850606283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,12288,0.045151111152437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,16384,0.13038933277130127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,8192,0.036200000180138484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,51200,0.39177600542704266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,10240,0.03985155622164408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,7168,0.031494223409228854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,12288,0.09897955258687337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,6144,0.026330666409598455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,8192,0.029096000724368628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,10240,0.09072355429331462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,7168,0.02609066665172577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,5120,0.021949332621362474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,4096,0.01834755473666721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,8192,0.0672524438963996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,7168,0.05917511383692423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,6144,0.05145333210627238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,3584,0.016557332542207506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,4096,0.01755644381046295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,2048,0.012574222352769641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,4096,0.03645688957638211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,3584,0.03252266512976752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,2560,0.012700444294346703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,3072,0.015108444624476962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,3072,0.02860266632503933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,2048,0.01107377807299296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,2560,0.013817778064144982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,2560,0.024890666206677754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,1536,0.009124444590674506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,2048,0.012646222280131446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,2048,0.021201777789327834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,1024,0.007084444165229797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,1536,0.011666666302416058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,1536,0.01757777730623881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,6144,0.022936888866954382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,1024,0.010493333141009012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,1024,0.013463999662134381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,5120,0.020198222663667467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,512,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,768,0.010212444596820408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,3072,0.01463822192615933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,256,0.0041982221106688184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,768,0.011718221836619906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,512,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,128,0.003524444583389494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,512,0.009918221996890174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,256,0.009132444030708736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,64,0.0035413333939181436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,256,0.008039111064540016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,32,0.003544888976547453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,128,0.009000000026490953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,128,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,64,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,65536,0.21243645085228813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,32,0.007376888559924231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,51200,0.15605778164333767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,65536,0.1625724501079983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,16384,0.051741331815719604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,768,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,65536,0.42025065422058105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,12288,0.03939022289382087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,51200,0.32980089717441136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,16384,0.04555022385385302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,16384,0.10978577534357707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,3584,0.016247999336984422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,12288,0.036285334163241915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,8192,0.023806222611003455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,12288,0.08364622460471259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,10240,0.03234400020705329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,7168,0.01941066649225023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,10240,0.07046133279800415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,8192,0.022295110755496558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,6144,0.017993777990341187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,8192,0.056755556000603564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,7168,0.02053600052992503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,5120,0.014701333310869006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,7168,0.05031288994683159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,6144,0.01796800063716041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,4096,0.012136889000733694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,6144,0.043862223625183105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,5120,0.01591822173860338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,3584,0.010775999890433418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,5120,0.037343111303117536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,4096,0.014160888062583076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,4096,0.03096888793839349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,3072,0.009668444593747457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,3584,0.013425778183672162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,5120,0.04333244429694282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,2560,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,3584,0.027813331948386297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,2048,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,3072,0.02471911079353756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,2560,0.011667555405033959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,2048,0.011011555790901184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,2560,0.021407999926143225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,2048,0.01828977796766493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,1024,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,1536,0.010386666489972008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,768,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,1536,0.015230221880806817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,1024,0.009369778136412302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,512,0.0038906666967603895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,1024,0.011892444557613797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,768,0.00903911143541336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,256,0.0032088889016045462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,512,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,512,0.008728000024954477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,128,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,256,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,256,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,64,0.002599999929467837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,128,0.008307555483447181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,3072,0.012437333663304647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,32,0.002891555635465516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,10240,0.03282666537496779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,1536,0.006205333189831839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,64,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,65536,0.1258151133855184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,32,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,51200,0.10388355784946018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,16384,0.038314667012956403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,65536,0.16302755143907335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,51200,0.12910044193267822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,65536,0.3425439993540446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,51200,0.26731377177768284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,768,0.010101333260536194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,16384,0.09049777852164374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,10240,0.021164443757798936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,12288,0.037124445041020714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,12288,0.06793689065509372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,51200,0.12957421938578287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,10240,0.03233511249224345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,7168,0.012736000120639801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,10240,0.056976000467936196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,8192,0.023181334137916565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,6144,0.011384888655609555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,128,0.006680888848172293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,8192,0.04639200038380093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,7168,0.020322667227851022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,7168,0.04114222195413377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,6144,0.018220444520314533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,4096,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,6144,0.03602666656176249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,5120,0.016198222835858662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,3584,0.007845333053006066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,5120,0.030863109562132094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,4096,0.014375110467274984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,3072,0.007142222589916653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,4096,0.025455110602908667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,3584,0.013462222284740873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,12288,0.028912001185946997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,2560,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,3072,0.012418666647540199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,3072,0.02032088902261522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,2048,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,8192,0.014692443940374585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,2560,0.01178311142656538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,2560,0.017866666118303936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,1536,0.004573333180612988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,2048,0.011056000159846412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,1024,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,2048,0.015493333339691162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,5120,0.009862222605281407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,768,0.003495111233658261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,1024,0.009463110731707679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,1024,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,512,0.0031600000543726813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,768,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,768,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,256,0.0027955555253558685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,512,0.008642666869693333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,512,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,16384,0.04485955503251818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,3584,0.02296177711751726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,256,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,64,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,256,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,32,0.0025235555238193935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,128,0.008278222547637092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,128,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,65536,0.09900444746017456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,64,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,32,0.005882666756709416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,51200,0.079821334944831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,1536,0.010312889185216691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,1536,0.012950222525331708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,65536,0.16338311301337347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,65536,0.2994648880428738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,12288,0.02163733376397027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,16384,0.046738667620552905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,51200,0.23515110545688203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,10240,0.014767110347747803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,16384,0.07871111234029134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,12288,0.03757244348526001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,8192,0.013156444662147097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,12288,0.0598506662580702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,10240,0.0326533317565918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,7168,0.01184177812602785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,10240,0.050436443752712674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,8192,0.022914666268560622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,6144,0.010311110979980892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,7168,0.020623111062579684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,8192,0.041166222757763334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,5120,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,7168,0.03651377889845107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,6144,0.0196115556690428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,4096,0.0074933336840735535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,5120,0.016865778300497267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,6144,0.03173777792188857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,3584,0.0074177781740824384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,5120,0.027091556125217017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,4096,0.014581332604090372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,3072,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,16384,0.030215998490651447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,4096,0.022689777943823073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,3584,0.014142221874660917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,2560,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,2048,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,51200,0.13053066200680202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,3584,0.020636444290479023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,3072,0.013064888616402945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,1536,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,2560,0.012080888781282636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,1024,0.003526222374704149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,3072,0.018557333283954196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,2048,0.011350221931934357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,2560,0.016252444850073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,768,0.00319377767542998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,1536,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,2048,0.014147554834683737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,1024,0.0094373333785269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,1536,0.01205599970287747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,256,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,768,0.00908622228437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,1024,0.009686222506894005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,128,0.0025128889828920364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,768,0.008473777936564552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,256,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,512,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,64,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,256,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,32,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,128,0.008108444511890411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,128,0.005380444642570283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,65536,0.09423377778795029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,51200,0.07534044318728976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,64,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,32,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,16384,0.02676711148685879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,12288,0.01812977757718828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,10240,0.013682666752073499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,65536,0.30042576789855957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,16384,0.07861422167883979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,12288,0.05951199928919474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,8192,0.011503111157152386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,7168,0.011047110789351992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,512,0.0028159999185138275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,10240,0.05043377810054355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,6144,0.009647111097971598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,128,0.002481777841846148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,5120,0.008296888735559251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,8192,0.04088800152142843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,4096,0.007214222517278459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,512,0.008628444539176093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,3584,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,7168,0.03607199920548333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,3072,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,2560,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,6144,0.03147822287347581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,5120,0.027112000518374976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,2048,0.004577777865860197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,4096,0.022682666778564453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,1536,0.00388711111413108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,3584,0.020463999774720933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,1024,0.0032275555034478507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,3072,0.01834133267402649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,768,0.0029200000895394217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,512,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,2560,0.016154666741689045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,256,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,2048,0.014137778017255994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,1536,0.011808888779746162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,128,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,64,0.0021111110432280437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,768,0.008360000120268928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,512,0.007207999626795451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,32,0.0021688888470331826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,51200,0.2349466747707791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,65536,0.08935644229253133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,128,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,64,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,51200,0.07118488682640924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,32,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,16384,0.026300445199012756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,12288,0.017222222354676988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,10240,0.011426666544543372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,65536,0.30015378528171116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,16384,0.07861600319544475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,8192,0.010203555226325989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,51200,0.2343813313378228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,12288,0.05975466966629028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,7168,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,6144,0.00850222259759903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,10240,0.050256000624762646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,8192,0.04097422295146518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,7168,0.03617599937650893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,6144,0.03153422143724229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,4096,0.009731555150614845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,3584,0.009354666703277165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,5120,0.027037334111001756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,1024,0.009441778063774109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,2560,0.006996444529957241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,4096,0.022654222117529974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,256,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,3584,0.02102311121092902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,2048,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,1536,0.0038480001191298165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,3072,0.018598222070270114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,1024,0.0031964443624019623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,768,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,2560,0.01649599936273363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,512,0.002663111107216941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,2048,0.014375999569892883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,256,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,1536,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,128,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,64,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,1024,0.009682666924264696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,32,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,768,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,512,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,256,0.00563822231358952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,128,0.005320000151793162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,5120,0.012270222107569376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,32,0.005201777650250329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,3072,0.008440000315507254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,64,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,16384,1.607428444756402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,16384,4.915712992350261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,16384,5.824277242024739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,16384,5.301288180881076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,51200,6.593481699625651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,12288,4.520576053195529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,12288,1.2673537996080186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,12288,4.0176268683539496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,10240,1.05258666144477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,12288,3.781195746527778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,10240,3.8531388176812063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,10240,3.369628482394748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,8192,0.8493430879380969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,10240,3.0598968929714627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,8192,2.809724383884006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,8192,2.6140533023410373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,8192,2.504430135091146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,7168,2.673086166381836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,51200,15.927251180013021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,7168,2.213583204481337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,7168,0.7399519814385308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,6144,0.6340666876898872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,7168,2.185420354207357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,6144,1.9601785871717665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,6144,1.949984868367513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,5120,1.8459421793619792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,5120,0.5295839839511448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,6144,1.846567153930664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,5120,1.6435609393649633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,4096,1.259022182888455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,4096,0.4360640048980713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,5120,1.5953564114040797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,3584,0.3930995464324951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,3584,1.0871715545654297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,4096,1.4125617345174153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,4096,1.2461537255181205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,3584,1.1701617770724826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,3072,0.3485537899865045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,3584,1.1272000206841362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,3072,0.9472204844156901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,3072,0.9805342356363932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,2560,0.7988524436950684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,3072,0.9601430892944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,2048,0.6040835380554199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,2560,0.8471146689520942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,2560,0.7952755292256674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,51200,16.98566182454427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,2048,0.26369955804612905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,1536,0.47937599817911786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,2048,0.9402355617947049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,2048,0.6339688830905491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,2560,0.30304710070292157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,1536,0.20577422777811685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,1024,0.3399466673533122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,1536,0.5545706748962402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,768,0.2733093367682563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,1024,0.16509155432383218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,1024,0.3935999870300293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,512,0.2123119963539971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,768,0.15294578340318468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,768,0.4327182239956326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,768,0.29683645566304523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,256,0.17244533697764078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,512,0.262140432993571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,256,0.11049866676330566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,256,0.23524888356526694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,512,0.2442071172926161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,128,0.10687466462453206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,51200,24.601021660698787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,256,0.1887200011147393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,64,0.10720622539520264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,128,0.1092462208535936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,32,0.10674577951431274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,128,0.22345333629184297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,1536,0.5411982006496853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,1024,0.36851199467976886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,128,0.14138844278123644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,64,0.22966578271653917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,512,0.13857244120703802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,32,0.2342595524258084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,51200,5.299909379747179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,65536,6.845409817165798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,16384,4.449115753173828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,51200,12.269002278645834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,16384,5.068690829806857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,16384,1.3220880296495225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,65536,16.081375122070312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,12288,3.3077884250217013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,16384,4.166920132107205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,12288,1.031344837612576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,12288,3.5322977701822915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,51200,15.55094740125868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,10240,3.081951141357422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,51200,20.004835340711804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,12288,2.9616932339138455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,10240,0.9087333679199219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,10240,2.8929369184705944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,10240,2.5024053785536027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,8192,0.6327599949306911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,8192,2.216628392537435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,8192,2.185477362738715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,65536,17.44295925564236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,8192,2.0022186703152127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,7168,0.5771679878234863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,7168,1.8362301720513239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,6144,1.5571218066745336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,6144,0.50800535413954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,65536,24.439261542426213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,7168,2.023909250895182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,7168,1.6978933546278212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,6144,1.4949412875705297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,5120,1.2695333692762587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,5120,0.42025065422058105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,6144,1.463645299275716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,4096,0.8592338032192655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,4096,0.3474577797783746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,5120,1.2497377395629883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,4096,1.0306658214992948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,5120,1.22998841603597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,3584,0.9795359505547417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,3584,0.3128746615515815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,4096,0.9539723926120334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,3072,0.2720239957173665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,3584,0.9008426666259766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,3072,0.6961778004964193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,3584,0.8083590931362576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,2560,0.6729404661390516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,3072,0.7852169142829047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,2560,0.25301067034403485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,3072,0.7357946501837836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,2048,0.5627662340799967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,2048,0.20165333482954237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,2560,0.6645849015977647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,2560,0.5948008961147732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,1536,0.3757217725118001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,2048,0.5684435632493761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,2048,0.49332888921101886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,1536,0.16131288475460476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,1024,0.26257777214050293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,768,0.2141351170010037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,1536,0.5572320090399848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,1536,0.38466665479871964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,512,0.15131378173828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,768,0.12126489480336507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,768,0.23394489288330078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,512,0.10559466812345718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,256,0.10213866498735215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,512,0.2064595619837443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,512,0.19011378288269043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,256,0.0888213316599528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,256,0.18750400013393828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,256,0.14795645078023276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,1024,0.13664889335632324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,64,0.07696800099478827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,128,0.17638222376505533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,1024,0.4125208854675293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,128,0.08734844128290813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,32,0.07474844323264228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,128,0.1127084493637085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,768,0.32350312338935006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,64,0.18091821670532227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,32,0.19006133079528809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,128,0.08766755792829727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,1024,0.2879626750946045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,65536,2.131574206882053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,51200,1.588414192199707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,16384,1.5112951066758897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,51200,4.023386637369792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,65536,4.995143890380859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,16384,0.4121653238932292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,16384,1.6241501702202692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,12288,1.184973292880588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,16384,1.2923866907755535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,12288,1.1090062459309895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,51200,4.475446065266927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,12288,0.9598755306667752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,12288,0.32720621426900226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,51200,6.588656107584636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,10240,0.26173777050442165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,10240,0.9321137534247504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,65536,5.968824174669053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,10240,0.8646373218960233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,8192,0.6371191342671713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,8192,0.2651982307434082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,10240,0.7355786429511176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,8192,0.6733902295430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,7168,0.18987466229332817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,7168,0.6494133207533095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,65536,8.312840779622396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,8192,0.5816462304857042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,7168,0.5811955663892958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,6144,0.4864311218261719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,6144,0.20444711049397787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,7168,0.5388444264729818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,5120,0.41886401176452637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,6144,0.48897420035468203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,5120,0.18220888243781197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,5120,0.442214224073622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,4096,0.38475465774536133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,4096,0.12067911359998916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,6144,0.444992012447781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,5120,0.3794737656911214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,4096,0.33457599745856387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,3584,0.3456097708808051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,3072,0.23411289850870767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,4096,0.36337688234117294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,3584,0.29251821835835773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,2560,0.20216088824801973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,3072,0.12269511487748887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,3072,0.2626897758907742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,2048,0.17948355939653185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,3584,0.29961956871880424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,3072,0.24135377671983507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,2560,0.22142310937245688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,1536,0.1359564463297526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,2048,0.17710044648912218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,2560,0.19976088735792372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,1536,0.0571991138988071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,2048,0.17469688256581625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,1536,0.14295821719699436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,3584,0.10796799924638535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,1536,0.13413333892822266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,768,0.0688213308652242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,1024,0.04780800143877665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,1024,0.10429244571261936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,2560,0.09262488947974311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,768,0.08380799823337132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,512,0.05522400140762329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,768,0.04365866714053684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,2048,0.06972710953818427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,768,0.08693599700927734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,256,0.03437422381507026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,512,0.03968000080850389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,512,0.07135288582907783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,128,0.029860443539089624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,256,0.03546222382121616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,512,0.07048266463809542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,256,0.0640159977806939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,1024,0.0850951141781277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,64,0.025038222471872967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,32,0.025382222400771245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,256,0.05556533071729872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,128,0.03522577881813049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,1024,0.10731289121839736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,64,0.062011553181542285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,128,0.04275111026234097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,32,0.06361244122187297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,128,0.06076622009277344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,65536,1.4951066970825195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,51200,1.19904359181722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,51200,3.171200858222114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,16384,1.2649830712212455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,65536,3.939312828911675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,16384,0.3330044481489393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,16384,1.1934827168782551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,12288,0.8161866929796008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,16384,1.008335961235894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,12288,0.8400906456841363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,12288,0.2664959960513645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,51200,3.4371244642469616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,12288,0.7049022250705295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,10240,0.7317991256713867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,10240,0.23104445139567056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,65536,4.450631035698785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,51200,5.701363457573785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,10240,0.6570409138997396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,65536,6.113459693060981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,8192,0.5422222349378798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,10240,0.5729937553405762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,7168,0.4802604251437717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,8192,0.6299929089016384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,7168,0.44491110907660586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,8192,0.4577013121710883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,7168,0.39874932501051163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,6144,0.1297066609064738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,5120,0.29696088367038304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,6144,0.5624684227837456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,5120,0.11138933234744602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,6144,0.3484728866153293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,4096,0.24259821573893228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,5120,0.3563226593865289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,8192,0.18498310777876112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,7168,0.17523733774820963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,4096,0.09712888797124226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,3584,0.20431555642022026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,4096,0.2571644518110487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,6144,0.36678844028049046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,3072,0.18302044603559706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,3584,0.08230044444402058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,3584,0.2204444408416748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,4096,0.24051645067003039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,2560,0.14894844426049128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,3072,0.07988888687557645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,3584,0.2095866600672404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,3072,0.19174310896131727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,2048,0.11973688337537979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,2560,0.06552888949712117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,2560,0.1643528938293457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,3072,0.18375466929541695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,2048,0.057222220632765025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,2560,0.15804444419013128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,2048,0.1352586613761054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,1536,0.04571288824081421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,1024,0.07502488957511054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,5120,0.28743023342556423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,1024,0.03890044490496317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,768,0.05599021911621094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,1024,0.08160889148712158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,1536,0.10841955741246541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,512,0.038535111480289035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,768,0.035360889302359685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,1024,0.0816355546315511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,768,0.06539289156595866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,256,0.02598222262329525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,512,0.03268444538116455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,512,0.05564710828993055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,768,0.06904266940222846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,1536,0.09261511431799994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,256,0.029458665185504492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,512,0.05626222160127428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,256,0.049766222635904946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,256,0.044848889112472534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,1536,0.10756533013449775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,128,0.02906844351026747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,64,0.018755555152893066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,128,0.04714577727847629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,32,0.019038221902317472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,2048,0.13208533657921687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,128,0.03458755546145969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,64,0.04797955685191684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,32,0.04875377813975016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,128,0.020958221620983545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,65536,1.3097537358601887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,51200,0.9703867170545789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,16384,0.9831110636393229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,51200,2.7567520141601562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,65536,3.282760832044813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,16384,0.27157333162095815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,16384,1.0810284084743924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,12288,0.701814227634006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,16384,0.8743822309705945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,51200,4.044284396701388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,12288,0.6710186534457736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,12288,0.2036257717344496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,51200,2.8994034661187063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,10240,0.17414132754007974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,12288,0.6116524272494847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,10240,0.5983991093105739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,65536,3.7667770385742188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,65536,5.012881808810763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,10240,0.5715137587653266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,8192,0.4402044349246555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,8192,0.17814489205678305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,10240,0.4889369010925293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,7168,0.39333510398864746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,8192,0.4330480098724365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,7168,0.14602222707536486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,8192,0.3965146541595459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,7168,0.4401422341664632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,6144,0.38585689332750106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,5120,0.2683315542009142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,7168,0.34000176853603786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,5120,0.10871111022101508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,4096,0.1940364440282186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,5120,0.27180443869696724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,6144,0.3475964334275987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,5120,0.2506062189737956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,3584,0.17370222674475777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,4096,0.21477244959937203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,3072,0.14570132891337076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,3584,0.07170844740337796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,4096,0.21768711672888863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,3584,0.19331110848320854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,2560,0.12716978126102024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,3072,0.06325599882337782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,3584,0.1803955501980252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,6144,0.11038400067223443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,3072,0.16340355078379312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,2048,0.10379022359848022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,6144,0.33519644207424587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,2560,0.05554844604598152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,3072,0.15962044397989908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,2560,0.13974044058057997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,1536,0.07778844568464491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,2048,0.04747555653254191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,2048,0.11536978350745307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,2560,0.13585510518815783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,1024,0.05897777610354953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,1536,0.0391813317934672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,1536,0.09155911207199097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,4096,0.07994311385684542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,2048,0.11482844087812637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,1024,0.07031910949283175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,512,0.03051911128891839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,1536,0.09291555484135945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,768,0.031503111124038696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,768,0.05624800258212619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,1024,0.07139466868506537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,256,0.02313599983851115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,512,0.028728889094458684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,768,0.06089422437879774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,512,0.0476435555352105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,128,0.02010311186313629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,256,0.026132444540659588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,256,0.0424124432934655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,512,0.05018844538264804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,64,0.016511999898486666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,32,0.01787911189926995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,128,0.02546399997340308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,128,0.040404445595211454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,256,0.03987911012437608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,64,0.04110044572088454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,128,0.0302782224284278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,32,0.041950222518708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,768,0.04467644294102987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,1024,0.03405155407057868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,65536,1.210209740532769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,51200,2.3603795369466147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,65536,2.62320793999566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,51200,0.9156817330254449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,16384,0.7283786667717828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,51200,2.3482240041097007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,51200,3.0718701680501304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,16384,0.257264879014757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,16384,0.8007831043667264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,16384,0.6615031030442979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,12288,0.1831244495179918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,65536,3.0218853420681424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,12288,0.5683804617987739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,65536,4.097070058186849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,10240,0.43985867500305176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,10240,0.1660391092300415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,12288,0.5689786805046929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,8192,0.3329075442420112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,12288,0.479996469285753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,8192,0.11858756012386745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,8192,0.36673688888549805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,7168,0.3217537668016222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,10240,0.3964248763190375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,7168,0.10707022084130181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,6144,0.22459112273322213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,7168,0.3554862340291341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,8192,0.35770222875807023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,5120,0.19696711169348824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,6144,0.09455644422107273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,6144,0.3113937907748752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,7168,0.28521066241794163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,4096,0.16137866179148355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,5120,0.09050933519999187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,5120,0.2263911035325792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,6144,0.24330933888753256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,3584,0.1355839967727661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,4096,0.07164533270729913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,5120,0.21265599462721083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,4096,0.18308533562554252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,3072,0.11961511770884196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,10240,0.47106843524509007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,3584,0.06614222129185994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,4096,0.1748302247789171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,3584,0.1542311112085978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,2560,0.11266933547125922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,3072,0.05398933423890007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,3584,0.14977422025468615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,3072,0.13470489448971218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,2048,0.08815822336408828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,2560,0.046929776668548584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,1536,0.0706168876753913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,2560,0.12431466579437256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,3072,0.13337243927849665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,2048,0.04015466570854187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,1024,0.04553066690762838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,2048,0.09576533238093059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,2560,0.11299110783471002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,1536,0.07544889052708943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,1536,0.03477511141035292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,768,0.03611022233963013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,2048,0.09580000241597493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,1024,0.028524445162879095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,1024,0.057461334599388965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,512,0.026176000634829204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,768,0.025640888346566096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,256,0.017872000734011333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,768,0.04606577754020691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,512,0.023224888576401606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,1024,0.06113600068622165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,128,0.014399111270904541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,512,0.038338667816585965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,768,0.05234488844871521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,256,0.020599111914634705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,64,0.013087111214796701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,512,0.042860445049073964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,32,0.013545778062608508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,128,0.020213334096802604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,256,0.033081776565975614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,64,0.031790223386552595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,128,0.02557955516709222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,32,0.032124443186653986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,1536,0.08160444762971666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,256,0.03348355491956075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,128,0.03154666556252374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,65536,0.9617172876993815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,51200,0.793168862660726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,65536,2.503357357449002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,16384,0.8525982432895236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,51200,2.3141119215223527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,16384,0.23559821976555717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,12288,0.5579048792521158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,16384,0.7958790991041395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,51200,2.747300465901693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,12288,0.4927822219000922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,12288,0.1593244473139445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,10240,0.3568284511566162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,51200,2.1441894107394748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,65536,4.037677341037326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,12288,0.45153066847059464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,8192,0.29533688227335614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,10240,0.14764532777998182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,10240,0.38574934005737305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,10240,0.3674160109625922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,8192,0.31461599138047963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,16384,0.6081902186075846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,8192,0.31334132618374294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,65536,2.796640819973416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,7168,0.26983022689819336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,6144,0.22889955838521323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,7168,0.26032887564765084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,5120,0.1792008876800537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,6144,0.22955645455254448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,6144,0.2245537704891629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,5120,0.07260000043445163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,8192,0.10671377844280666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,4096,0.13511021931966147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,7168,0.2742391162448459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,5120,0.19447199503580728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,3584,0.11885066827138265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,4096,0.060772445466783315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,5120,0.19188888867696127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,4096,0.15396800306108263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,3072,0.10193510850270589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,3584,0.0543431109852261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,3584,0.13662400510576037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,4096,0.15608621968163386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,2560,0.0915351112683614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,3072,0.048306667142444186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,3584,0.13890133963690862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,2048,0.07534044318728976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,2560,0.10288444492552017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,3072,0.12331377135382758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,2560,0.10504088799158733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,2048,0.036406222316953875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,2048,0.08495733473036025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,1536,0.05265600151485867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,2048,0.088992887073093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,1536,0.030088000827365454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,1024,0.04462488823466831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,1536,0.0675813357035319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,7168,0.09555644459194607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,1536,0.0720391141043769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,768,0.033352000845803156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,1024,0.0509057773484124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,3072,0.11937067243787979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,768,0.0232924438185162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,512,0.02630311085118188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,768,0.041843556695514254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,2560,0.042576889197031655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,6144,0.0836355553732978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,256,0.01993955506218804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,768,0.047888888253106006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,512,0.02069511181778378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,512,0.0340008901225196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,128,0.01465777721669939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,256,0.018943111101786297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,256,0.02934044599533081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,512,0.03938755393028259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,64,0.01241422196229299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,32,0.01275466630856196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,256,0.03090755475891961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,128,0.0185208883550432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,128,0.028040889236662123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,128,0.02330933345688714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,64,0.028395556741290625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,32,0.02812266680929396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,1024,0.025647999511824712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,1024,0.056142224205864795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,65536,0.8375395668877496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,51200,0.6626710891723633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,51200,1.5416026645236547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,16384,0.6831173366970487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,65536,2.158743116590712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,16384,0.18860177199045816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,16384,0.6304826736450195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,12288,0.39786489804585773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,51200,1.8749287923177083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,16384,0.5359422365824381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,12288,0.14895199404822454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,65536,3.1327156490749783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,65536,2.3939208984375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,10240,0.3599075476328532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,12288,0.38841954867045086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,10240,0.11719377835591634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,10240,0.3701146708594428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,8192,0.2943342261844211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,8192,0.0932355523109436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,7168,0.21988177299499512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,10240,0.32826222313774955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,8192,0.35683290163675946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,7168,0.08417599731021458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,8192,0.26231111420525444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,6144,0.20399999618530273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,7168,0.31700176662868923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,51200,2.367788526746962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,5120,0.16549955474005804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,6144,0.0737537807888455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,6144,0.20916888448927137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,7168,0.22681866751776802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,4096,0.1317395501666599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,5120,0.06930133369233873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,5120,0.1878417730331421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,3584,0.10831644137700398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,6144,0.19843555821312797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,4096,0.055043554968304105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,5120,0.1670231156879001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,4096,0.136698669857449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,3072,0.09943555461035834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,3584,0.04993155598640442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,4096,0.13833155896928576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,3584,0.12077511681450738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,3072,0.043410665459103055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,3072,0.10502666897243923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,3584,0.12335999806722005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,2048,0.07105511426925659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,2560,0.037830223639806114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,3072,0.10859644412994385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,2560,0.09036533037821452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,1536,0.047172443734274976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,2048,0.0318195554945204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,2560,0.09500710831748115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,2048,0.07492978043026395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,12288,0.43083733982510036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,1024,0.03593333231078254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,1536,0.05895377529991997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,1536,0.02693066663212246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,2048,0.08004533582263522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,768,0.029821332958009508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,1024,0.022456889351209004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,1536,0.06427822510401408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,512,0.021454221672481958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,1024,0.050372444921069674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,2560,0.07772444354163276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,768,0.042985777060190834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,256,0.01512977812025282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,512,0.01846311158604092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,512,0.03561688794030084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,256,0.02573777735233307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,128,0.012455999851226807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,256,0.017159110969967313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,256,0.027791112661361694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,64,0.010759111079904767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,128,0.016955556141005624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,128,0.024547555380397375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,32,0.011503111157152386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,1024,0.04504088891877068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,128,0.020983111527231004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,64,0.02473955518669552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,768,0.020439111524158053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,32,0.024375999967257183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,768,0.03710133168432448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,512,0.02966133422321743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,65536,0.7036106851365832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,51200,0.5537101957533095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,65536,1.631694263882107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,16384,0.5040559768676758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,51200,1.4865404764811199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,16384,0.19759644402398002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,12288,0.3833173380957709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,51200,1.5636284086439345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,51200,2.101365407307943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,16384,0.4637600051032172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,12288,0.12325333224402533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,65536,2.6021804809570312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,12288,0.4165324370066325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,10240,0.29441867934332955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,65536,2.080916510687934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,10240,0.10719022485944961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,8192,0.23153334193759492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,12288,0.3481173250410292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,10240,0.3110311031341553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,8192,0.08318666617075603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,8192,0.26633956697252065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,10240,0.2801297770606147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,7168,0.21189333332909477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,16384,0.5308924251132542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,8192,0.23620798852708605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,7168,0.2029351128472222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,5120,0.1344568861855401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,6144,0.06388444370693631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,6144,0.17515644762251112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,7168,0.19852977328830293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,4096,0.10353600316577488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,6144,0.17678044901953804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,5120,0.05506844321886698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,5120,0.14635645018683538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,3584,0.09287999735938178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,4096,0.0469742218653361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,4096,0.11663288540310329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,5120,0.14574844307369655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,3072,0.07820800277921888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,3584,0.04133244355519613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,3584,0.10355377859539455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,4096,0.12279733022054036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,2560,0.060398221015930176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,3072,0.03674044542842441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,3584,0.10752444134818183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,3072,0.09073333607779609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,2048,0.058971557352277965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,2560,0.03293244375122918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,7168,0.07401955789989896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,2560,0.07794133159849378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,3072,0.09456888834635417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,1536,0.04254577888382805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,2048,0.02757866680622101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,2560,0.08132621977064344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,2048,0.06419555346171062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,1024,0.02914488977856106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,1536,0.051544000705083214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,1536,0.022914666268560622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,768,0.02353422178162469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,2048,0.06938755512237549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,1024,0.019304000669055514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,1536,0.05757155683305529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,1024,0.03888622257444594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,512,0.01703377730316586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,768,0.017882666654056974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,1024,0.04400711258252462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,768,0.032243556446499295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,256,0.01239377756913503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,512,0.025674665967623394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,512,0.016255999604860943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,768,0.03800444470511542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,128,0.01014755583471722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,256,0.015164444843928019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,256,0.02230311102337307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,64,0.008978666530715095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,6144,0.18113422393798828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,128,0.021474666065639917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,128,0.015160888433456421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,128,0.018740443719757926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,64,0.021345777644051447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,32,0.021271111236678228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,65536,0.5800062285529243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,512,0.03193333413865831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,65536,1.2168773015340169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,256,0.024148444334665935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,51200,1.0782000223795574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,16384,0.3346417744954427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,51200,0.4452142185635037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,16384,0.47915108998616535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,16384,0.1333813269933065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,65536,2.109823015001085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,65536,1.7332942750718858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,12288,0.2629244327545166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,16384,0.3749573230743408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,32,0.010683555569913654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,12288,0.10345155662960476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,10240,0.24561778704325357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,51200,1.298249774509006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,51200,1.6573030683729384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,8192,0.20336800151401094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,10240,0.09069510963228013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,10240,0.23955376942952475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,12288,0.2808604505327013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,7168,0.15993777910868326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,8192,0.08914044168260361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,10240,0.2304017808702257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,8192,0.24043554729885527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,7168,0.06263377931382921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,6144,0.15830043951670328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,7168,0.16758755842844644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,8192,0.201110217306349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,5120,0.12802844577365452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,6144,0.05575022432539198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,7168,0.16609244876437718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,4096,0.08508888880411784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,5120,0.12187822659810384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,5120,0.05289333396487766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,6144,0.14751911163330078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,3584,0.08014933268229167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,4096,0.038551112016042076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,4096,0.09751733144124348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,5120,0.12457511160108779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,3072,0.07019733058081733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,3584,0.03372800019052293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,4096,0.10263466835021973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,2560,0.05229155553711785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,3072,0.030307554536395605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,3584,0.09250311056772868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,2560,0.028282665544086035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,3072,0.08037510845396253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,2560,0.06570755773120456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,2048,0.043712887499067515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,2048,0.02398577829202016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,6144,0.14495733049180773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,2560,0.06929066446092394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,1536,0.0331911113527086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,2048,0.054302222198910184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,1024,0.023891554938422307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,1536,0.020495999190542433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,2048,0.05961155229144626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,1536,0.04334400097529093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,12288,0.29814222123887807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,768,0.018901333212852478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,1024,0.017398221625222098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,1024,0.032891554964913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,1536,0.04840799834993151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,512,0.014013333453072442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,3584,0.08692177799012925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,768,0.015895111693276297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,768,0.027284445034133062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,3072,0.07545777824189928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,256,0.010656889114114972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,512,0.014483556151390076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,512,0.02623822291692098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,256,0.013618666264745923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,256,0.01888444523016612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,128,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,256,0.022039999564488728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,64,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,128,0.013803555733627744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,32,0.008096000386608971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,128,0.018239999810854595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,128,0.017254221770498488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,64,0.018244443668259513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,32,0.017879999346203275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,1024,0.03751911057366265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,768,0.03266577919324239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,65536,0.5001440048217773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,65536,1.3617910808987086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,51200,0.9450719621446398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,51200,0.39197956191168887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,16384,0.35626310772365993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,65536,1.7078417672051325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,16384,0.13194933202531603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,512,0.021983999345037673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,51200,1.4192808998955622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,16384,0.391995562447442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,65536,1.5344479878743489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,12288,0.2572462293836805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,51200,1.1991572909884982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,16384,0.37775643666585285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,10240,0.20842133627997506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,12288,0.09950133164723714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,12288,0.26308978928460014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,8192,0.16163289546966553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,10240,0.2387324439154731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,12288,0.27294309933980304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,8192,0.18286311626434326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,10240,0.24393065770467123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,8192,0.19113777743445504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,7168,0.05817511346605089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,6144,0.11966133117675781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,7168,0.15190754996405706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,5120,0.08928000264697605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,7168,0.1631315549214681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,6144,0.050048000282711454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,6144,0.13016443782382542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,4096,0.08038666513231066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,6144,0.14209155241648355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,5120,0.10948799716101752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,10240,0.08375733428531223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,5120,0.12057244777679443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,4096,0.03506222367286682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,4096,0.08989689085218643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,3584,0.06520266665352716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,7168,0.14705777168273926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,3072,0.06007555458280775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,3584,0.03163377775086297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,4096,0.09910577535629272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,3584,0.07832800017462836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,2560,0.04715377754635281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,3072,0.028706666496064927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,3072,0.06892266538408068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,2048,0.03879822293917338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,3584,0.08839555581410725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,2560,0.025569778349664476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,3072,0.07981689108742608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,2560,0.0586035582754347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,1536,0.029078221983379785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,2048,0.021996445126003687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,2560,0.0682800014813741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,5120,0.042430222034454346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,1024,0.020294222566816542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,1536,0.01867644488811493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,1536,0.03957688808441162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,2048,0.057592001226213246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,768,0.01684266659948561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,1024,0.01594311164485084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,1536,0.04805244339836968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,1024,0.029341333442264136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,512,0.01294133315483729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,8192,0.06338666545020209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,768,0.01516711049609714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,768,0.024431111084090337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,1024,0.036650667587916054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,256,0.009334222310119206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,512,0.013855111267831592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,512,0.0195777780479855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,128,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,256,0.013059555656380124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,256,0.01720977822939555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,64,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,128,0.013064000341627332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,32,0.00889511075284746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,256,0.021319111188252766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,128,0.01649599936273363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,128,0.016882666283183627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,64,0.016173332929611206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,2048,0.04945422212282816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,32,0.016197333733240765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,65536,0.4398524496290419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,768,0.03068800104988946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,65536,1.0534719891018338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,512,0.02533866630660163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,51200,0.8805955780877007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,16384,0.2785617775387234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,51200,0.36905866199069554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,16384,0.33931377198961044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,65536,1.5297608905368383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,16384,0.10806222094429864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,12288,0.19498133659362793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,65536,1.4144781960381403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,16384,0.3404364320966933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,10240,0.17193510797288683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,12288,0.23736800087822807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,51200,1.2173955705430772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,51200,1.0536808437771268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,8192,0.13309778107537165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,10240,0.07197866837183635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,12288,0.24921955002678764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,10240,0.19737777445051405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,7168,0.12833689318762884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,8192,0.06134044461780124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,8192,0.15099910895029703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,6144,0.1031128896607293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,10240,0.2174453337987264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,7168,0.05190044310357836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,7168,0.13391910658942327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,5120,0.08541511164771186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,6144,0.045062220758861966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,6144,0.1145342191060384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,4096,0.06784266895718045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,5120,0.03818133473396301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,5120,0.09775022003385757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,6144,0.1294462283452352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,12288,0.08299911022186279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,3584,0.055946667989095054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,5120,0.11044177744123672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,4096,0.03167644474241469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,3072,0.04808800088034736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,4096,0.07897422048780653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,3584,0.02885955572128296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,3584,0.06940978103213839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,4096,0.09084177679485744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,2560,0.041125333971447416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,3072,0.026241777671708003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,3584,0.08124533626768324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,3072,0.06042666567696465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,2048,0.03387466735310025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,2560,0.023426666855812073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,3072,0.07198311222924127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,2560,0.05184355709287855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,8192,0.1724720001220703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,1536,0.02648266653219859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,2048,0.020643555455737643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,2048,0.04382755690150791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,2560,0.0613573326004876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,1024,0.019096889429622226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,7168,0.1491031116909451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,1536,0.01809155609872606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,768,0.014863111906581454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,1536,0.03545511100027297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,2048,0.05257244573699104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,1024,0.015505777464972602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,512,0.011439111497667102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,1024,0.026370666093296472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,1536,0.043559110826916166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,256,0.009741333623727163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,768,0.02233955595228407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,512,0.01313244468635983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,512,0.01796444422668881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,768,0.028283556302388508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,256,0.01239377756913503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,512,0.02369511127471924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,256,0.015454222758611044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,64,0.008060444560315874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,128,0.012406222522258759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,128,0.015088889333936902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,32,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,64,0.014465777410401238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,128,0.015969778100649517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,32,0.014565333724021912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,768,0.014482667048772177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,1024,0.03349333339267307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,65536,0.7764995363023547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,128,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,256,0.018887110882335238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,51200,0.5895528793334961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,51200,0.30885420905219185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,65536,1.1882248984442818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,16384,0.23511022991604277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,12288,0.17268889480166963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,16384,0.09922222296396892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,16384,0.26902222633361816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,51200,0.9036915037367078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,10240,0.15014666981167263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,12288,0.07562222083409627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,16384,0.29601510365804035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,65536,0.39203821288214785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,51200,0.9251182344224719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,12288,0.19818311267428926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,8192,0.11761421627468532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,10240,0.06386844317118327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,12288,0.22202132807837593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,10240,0.16381866402096218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,7168,0.10086844364802043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,8192,0.05069422059588962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,8192,0.13240444660186768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,10240,0.18533599376678467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,6144,0.08977244297663371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,7168,0.04570577873124016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,7168,0.11570933130052354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,65536,1.1572258207533095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,5120,0.07335555553436279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,6144,0.10162044233745998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,7168,0.1327306694454617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,4096,0.05600444475809733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,6144,0.11475200123257107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,5120,0.08553955290052627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,5120,0.09731733136706883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,4096,0.028543998797734577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,3584,0.04809244473775228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,4096,0.08149688773685031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,3584,0.061202665170033775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,3072,0.035331557194391884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,3584,0.025762667258580525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,3584,0.07238844368192884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,8192,0.15709511439005533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,2560,0.03386666708522373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,6144,0.03956266575389438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,3072,0.02313333253065745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,3072,0.053752889235814415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,2048,0.025843555728594463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,2560,0.02160266704029507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,2560,0.046322666936450534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,5120,0.033717334270477295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,3072,0.06456266509162055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,1536,0.021262221866183813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,2048,0.018578666779730055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,1024,0.01535555554760827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,2048,0.039072000318103366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,1536,0.0158942225906584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,4096,0.0687333345413208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,768,0.012251555919647217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,1536,0.03132799930042691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,2048,0.04750488864050972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,1024,0.013983110586802164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,1024,0.02380888826317257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,512,0.009368889033794403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,1536,0.03899822301334805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,768,0.013384888569513956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,768,0.0199937770764033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,1024,0.030140442980660334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,512,0.01221777747074763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,512,0.016163556112183463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,768,0.0258906665775511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,128,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,256,0.011595555477672152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,256,0.013785777820481194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,512,0.02126311096880171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,64,0.005531555662552516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,32,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,128,0.01165422217713462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,256,0.018239110708236694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,128,0.013232000172138214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,64,0.013075555364290873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,128,0.013960889644092984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,32,0.013192888763215808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,2560,0.056616001658969455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,65536,0.32649776670667857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,65536,0.6721839904785156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,256,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,51200,0.5462000105116103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,16384,0.1924462185965644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,51200,0.2565608819325765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,65536,0.9215075174967448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,51200,0.7218097580803765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,16384,0.09218844440248276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,12288,0.13872533374362522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,16384,0.21883911556667754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,10240,0.11670577526092529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,12288,0.07454133033752441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,65536,0.9981128904554578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,16384,0.2637520101335314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,51200,0.7693324618869357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,12288,0.169304887453715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,10240,0.05743022097481621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,10240,0.13947910732693142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,7168,0.08219910992516412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,12288,0.19554400444030762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,8192,0.11312354935540093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,6144,0.07013244761361016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,7168,0.038600001070234514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,8192,0.1329937775929769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,6144,0.034429334931903414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,7168,0.1177626715766059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,6144,0.08620533015992905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,5120,0.056835558679368764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,6144,0.10143555535210504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,5120,0.07315733035405476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,5120,0.02889510989189148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,4096,0.043659554587470166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,8192,0.09712888797124226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,5120,0.08697955475913154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,3584,0.03726666503482395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,4096,0.025221332907676697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,8192,0.04452266626887851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,4096,0.05811466773351034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,3072,0.035614222288131714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,3584,0.022848889231681824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,10240,0.15966755814022487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,3584,0.05264711048867968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,7168,0.09900710980097453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,2560,0.029748443100187514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,3072,0.020675554871559143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,3072,0.045979556110170156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,3584,0.0647982226477729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,2048,0.02311822275320689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,2560,0.01864266726705763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,2560,0.03977600071165297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,1536,0.017673777209387887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,3072,0.057722667853037514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,2048,0.016591111818949383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,2048,0.0335822237862481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,2560,0.05009777678383721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,1024,0.012808889150619507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,1536,0.014875556031862894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,2048,0.042457779248555504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,1536,0.027207111318906147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,768,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,1024,0.020654223031467862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,1024,0.013032889200581444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,1536,0.034958223501841225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,512,0.008630221916569604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,768,0.012142221960756512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,768,0.017291555802027386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,1024,0.027222222752041284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,4096,0.07274044222301908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,512,0.014109333356221518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,768,0.023193778263198003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,128,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,256,0.010703999963071613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,256,0.012096000214417776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,512,0.01926488843229082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,64,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,32,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,128,0.011908444265524546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,128,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,256,0.016232000456915963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,64,0.011376000112957425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,128,0.01311466677321328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,32,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,65536,0.44214489724900985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,65536,0.26844800843132866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,51200,0.4362195597754584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,256,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,16384,0.14389688438839382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,65536,0.745664013756646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,51200,0.2135253349939982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,51200,0.5865466859605577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,12288,0.10551466544469197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,16384,0.07088533374998304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,16384,0.181931561893887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,65536,0.8359866672092013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,10240,0.08886577685674031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,12288,0.1393466658062405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,12288,0.06002133422427707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,8192,0.07380711370044284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,51200,0.6501697964138454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,10240,0.052298665046691895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,10240,0.12589066558414036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,12288,0.16322933302985296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,7168,0.06153866979810926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,8192,0.03955022162861294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,8192,0.0947822199927436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,10240,0.13931555218166775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,6144,0.053688890404171415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,7168,0.034156444999906756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,8192,0.11451733112335205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,512,0.011357333097192975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,5120,0.042149331834581166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,6144,0.030894223186704848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,6144,0.07134755452473958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,4096,0.032297776805029974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,5120,0.060779558287726514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,5120,0.02622755534119076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,6144,0.08700266811582778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,3584,0.028928889168633357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,4096,0.02160444524553087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,5120,0.07471733623080783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,16384,0.22544889979892305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,3072,0.025980444418059453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,3584,0.01943911115328471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,4096,0.06338577800326876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,3072,0.01790577835506863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,3072,0.03918222255176968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,2560,0.022643556197484333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,7168,0.08654844760894775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,3072,0.049786667029062905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,2560,0.03436977664629618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,2560,0.016355555918481614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,7168,0.10352177752388848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,2048,0.019462222854296368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,2048,0.015831111205948722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,2560,0.04292710953288608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,1536,0.01314399970902337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,2048,0.028982222080230713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,4096,0.04974844389491611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,1024,0.010392000277837118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,1536,0.013271110753218332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,2048,0.035924444595972695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,3584,0.044715556833479136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,1536,0.023353776997990076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,768,0.008739555875460306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,1024,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,1024,0.017881777551439073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,512,0.006992888947327931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,3584,0.056122667259640165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,768,0.011051555474599203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,768,0.014837332897716098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,1024,0.02368800010946062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,512,0.010434666441546546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,768,0.020851555797788832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,512,0.012496888637542725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,128,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,256,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,512,0.018969777557584975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,64,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,128,0.009992889232105678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,128,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,32,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,256,0.015792888071801927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,64,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,128,0.011990221838156382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,32,0.00980088859796524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,65536,0.32290487819247776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,1536,0.030900445249345567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,51200,0.2408897876739502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,65536,0.21575466791788736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,256,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,16384,0.09779733419418335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,65536,0.562428421444363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,51200,0.1693377759721544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,51200,0.45383554034762913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,16384,0.05734755595525106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,256,0.010678222609890832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,12288,0.07011733452479045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,16384,0.1438657840092977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,10240,0.05948177973429362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,12288,0.045475555790795215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,16384,0.17926311492919922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,12288,0.10993511146969265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,51200,0.5308186742994521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,10240,0.04018400112787882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,8192,0.049698667393790356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,10240,0.09159466955396865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,7168,0.04046133491728041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,12288,0.13834933439890543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,8192,0.029806223180558946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,10240,0.11673777633243138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,6144,0.03319199879964193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,8192,0.07458311319351196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,7168,0.026226666238572862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,7168,0.06520089175966051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,5120,0.02834222382969326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,6144,0.023639111055268183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,8192,0.0972399976518419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,4096,0.023787554767396715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,7168,0.08481244246164958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,5120,0.02051911089155409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,5120,0.0477111107773251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,6144,0.07313599851396349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,3584,0.0211182220114602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,4096,0.017626666360431247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,4096,0.039841777748531766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,5120,0.0635040005048116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,3072,0.018624888526068795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,65536,0.6791875627305773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,3584,0.016324443949593436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,3584,0.035800000031789146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,2560,0.016226665841208566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,4096,0.05293600095642937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,3072,0.03153955605294969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,3072,0.015022221538755627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,3584,0.04684355523851183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,2560,0.013803555733627744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,2560,0.027062222361564636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,3072,0.04221777783499824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,1536,0.01106844428512785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,2048,0.012629333469602795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,2048,0.02288444505797492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,2560,0.035832888550228544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,1024,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,1536,0.011383111278216044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,1536,0.018982221682866413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,6144,0.05623822079764473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,1024,0.010412444670995077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,1024,0.014565333724021912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,1536,0.025754666990704004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,1024,0.020634666085243225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,768,0.010016889207892949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,512,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,768,0.012629333469602795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,768,0.01825066738658481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,512,0.010679110884666443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,256,0.004565333326657613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,2048,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,512,0.01583377851380242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,128,0.003887999802827835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,256,0.009007111191749573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,256,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,256,0.013772444592581855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,128,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,64,0.003510222252872255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,128,0.008627555436558193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,2048,0.03017688790957133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,32,0.0035413333939181436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,768,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,128,0.010667555862002902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,64,0.008267555799749162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,32,0.00815466625822915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,65536,0.30142490069071454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,51200,0.23136978679233125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,65536,0.2082915571000841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,16384,0.07302755779690213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,512,0.009320889082219865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,65536,0.5036835670471191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,51200,0.16495554976993138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,51200,0.3940862284766303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,12288,0.05501955416467455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,16384,0.05647110939025879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,16384,0.13042578432295057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,65536,0.6702764299180773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,12288,0.04509866568777296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,12288,0.09896800253126357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,51200,0.54712798860338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,8192,0.03496444556448195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,10240,0.0830577810605367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,12288,0.13793867164187962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,10240,0.03968266646067301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,7168,0.031086222993002996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,8192,0.029333333174387615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,10240,0.11637600262959798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,6144,0.02589955594804552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,7168,0.02560088866286808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,7168,0.05894666910171509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,8192,0.09530399905310737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,5120,0.02335111134582096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,7168,0.08356533447901408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,6144,0.022839110758569505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,6144,0.051348444488313466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,4096,0.01901244454913669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,5120,0.020308444897333782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,10240,0.046422223250071205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,5120,0.04355288876427544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,6144,0.07241600089603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,3584,0.01683822274208069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,4096,0.03647111190689935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,4096,0.017503110898865592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,5120,0.0620888868967692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,3584,0.016164445214801364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,3584,0.03261155552334256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,4096,0.052051554123560585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,2560,0.012883555557992725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,3072,0.014951111541854011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,3072,0.028483556376563177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,8192,0.0678079989221361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,2048,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,2560,0.013781333135233985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,2560,0.02516533268822564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,3072,0.04140977727042304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,1536,0.009023110899660323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,2048,0.021038222644064162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,2048,0.012664000193277994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,2560,0.03528977764977349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,1024,0.007246221933099959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,1536,0.011462222370836469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,1536,0.01752800080511305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,2048,0.030228445927302044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,768,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,1024,0.01036800030204985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,1024,0.013430222041077085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,16384,0.17734489176008436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,1536,0.02513866623242696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,3072,0.01479644411140018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,768,0.010050666828950247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,768,0.011477332976129321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,1024,0.020643555455737643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,256,0.0038906666967603895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,512,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,512,0.009745777481132084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,3584,0.04673155479960971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,128,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,768,0.017853332890404593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,256,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,512,0.015509333875444202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,256,0.008030222521887885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,128,0.008993777963850234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,32,0.0032231110251612137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,128,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,256,0.013456000222100152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,64,0.007371555599901412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,128,0.010728889041476779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,65536,0.193740447362264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,32,0.007332444190979004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,51200,0.15821866194407144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,65536,0.16087199581993952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,512,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,65536,0.41782400343153214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,16384,0.05346044566896227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,51200,0.12856711281670466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,51200,0.32848800553215873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,12288,0.04081155525313483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,16384,0.044907556639777295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,16384,0.10966310898462932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,65536,0.5288124614291722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,64,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,10240,0.034141331911087036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,12288,0.03598399957021078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,12288,0.08357777860429551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,51200,0.4893333117167155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,8192,0.024541333317756653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,10240,0.03217688865131802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,12288,0.10945955249998306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,8192,0.05724089013205635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,8192,0.02235288918018341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,10240,0.09407466650009155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,8192,0.07662933402591281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,7168,0.01980977753798167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,6144,0.018287110659811232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,7168,0.05053777827156914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,6144,0.017601778109868366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,7168,0.06715377834108141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,5120,0.015990222493807476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,6144,0.058840003278520375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,4096,0.013772444592581855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,5120,0.03716533382733663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,5120,0.015562666787041558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,10240,0.07022133138444689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,3584,0.012100444071822695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,4096,0.013776000175211163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,4096,0.03090488910675049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,7168,0.02125155594613817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,3072,0.010751999914646149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,3584,0.013055110971132914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,4096,0.04199022385809156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,3584,0.027815110153622095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,2560,0.00959822204377916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,3072,0.012434666355450949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,3072,0.024512888656722173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,3584,0.03746844331423441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,2048,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,2560,0.021432888176706102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,6144,0.04377866784731547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,3072,0.03325066632694668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,1536,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,2048,0.010776888993051318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,2048,0.018253333038753934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,2560,0.029024001624849107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,16384,0.1421440045038859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,1024,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,1536,0.010145777629481422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,1536,0.015154666370815702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,768,0.004566222015354368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,2048,0.025125333004527624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,5120,0.0505528880490197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,1024,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,1024,0.011958222422334882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,512,0.004173333446184794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,1536,0.02160088883505927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,768,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,1024,0.017539555827776592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,256,0.003499555504984326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,512,0.008380444513426887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,512,0.00868711123863856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,128,0.0028373334142896864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,256,0.008285333712895712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,512,0.013459555804729462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,256,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,64,0.0028355556229750314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,2560,0.011422221859296164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,128,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,32,0.0028328889360030494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,256,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,128,0.006657777975002925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,65536,0.12773066096835667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,64,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,128,0.009373333719041612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,32,0.00655733338660664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,51200,0.09531644317838882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,65536,0.1577813360426161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,65536,0.34099732504950625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,16384,0.03824888997607761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,768,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,51200,0.12671377923753527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,51200,0.26675465371873647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,12288,0.029234667619069416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,16384,0.04442666636572944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,768,0.015471110741297403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,65536,0.5105875333150228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,16384,0.08984177642398411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,10240,0.022044445077578228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,12288,0.06790933344099256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,12288,0.03546222382121616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,16384,0.13605599933200413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,8192,0.017326222525702584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,10240,0.05696711275312635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,10240,0.03151999910672506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,7168,0.015426667200194465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,12288,0.10711999734242757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,8192,0.021709332863489788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,8192,0.046448889705869884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,6144,0.013272888958454132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,7168,0.01920088960064782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,7168,0.041195554865731135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,8192,0.07396444347169664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,5120,0.011821333732869891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,6144,0.03603822324011061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,6144,0.017228444417317707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,5120,0.030989332331551447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,4096,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,6144,0.05673866801791721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,51200,0.40150221188863117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,5120,0.04840088884035746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,3584,0.009307555854320526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,4096,0.02537422213289473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,4096,0.013818666338920593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,3072,0.008207999997668797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,3584,0.013112000293201871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,4096,0.03997333182228936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,10240,0.09116355578104655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,3584,0.02275733318593767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,2560,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,3072,0.012318222059143914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,3584,0.0362702210744222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,3072,0.02033422225051456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,2048,0.006169777777459886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,2560,0.011730666789743634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,2560,0.01791733337773217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,7168,0.06493866443634033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,3072,0.03202666507826911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,2048,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,2048,0.015511110424995422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,2560,0.02809422214825948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,1024,0.004522666749027041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,5120,0.015526221858130561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,1536,0.010030222435792288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,1536,0.012998222476906247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,768,0.0038986665507157645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,2048,0.02427644365363651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,1024,0.009341333475377824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,512,0.003367111086845398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,1536,0.02058844433890449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,1024,0.01034044474363327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,768,0.008988444175985124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,256,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,768,0.008920888933870528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,1024,0.016753777861595154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,512,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,512,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,128,0.002493333278430833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,768,0.015136889285511441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,256,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,64,0.002479111154874166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,256,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,32,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,512,0.01313244468635983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,128,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,128,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,256,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,65536,0.09285244676801895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,64,0.005879999862776862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,51200,0.07494755585988362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,128,0.00865955568022198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,65536,0.15811378426022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,65536,0.30027378929985893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,16384,0.028210666444566514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,51200,0.12640621927049425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,51200,0.23501955138312447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,16384,0.04516177707248264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,16384,0.07845422294404772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,65536,0.5037048657735189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,51200,0.3967440128326416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,16384,0.1364977757136027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,10240,0.015862221519152325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,12288,0.059821334150102407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,12288,0.036473777559068464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,8192,0.013270222478442721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,10240,0.03264977865748935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,10240,0.050640887684292264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,12288,0.10630044672224258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,7168,0.011964444484975604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,32,0.005971555494599872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,8192,0.040955556763543025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,8192,0.022151110900772944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,10240,0.08997244305080837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,6144,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,7168,0.02016266683737437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,8192,0.07278044356240167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,6144,0.03174133433236016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,12288,0.022261333134439256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,7168,0.06433510780334473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,5120,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,6144,0.01814311080508762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,4096,0.007626666790909237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,5120,0.016206221448050607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,6144,0.056319110923343234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,5120,0.027328888575236004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,3584,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,4096,0.022681777675946552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,4096,0.014214222629865011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,5120,0.048122667604022555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,3072,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,3584,0.013448889056841532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,4096,0.03903466794225905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,3584,0.020609777834680345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,1536,0.005290666802061929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,2560,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,3072,0.0188080006175571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,3584,0.0351235568523407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,2048,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,2560,0.01180266671710544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,3072,0.03159911102718777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,2560,0.02767555581198798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,1536,0.004328000048796336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,2048,0.01421688828203413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,2048,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,2048,0.023544000254737005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,1024,0.0035573335157500375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,1536,0.012031111452314587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,1536,0.020311999652120803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,1024,0.009734222458468543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,768,0.0032151111712058387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,1024,0.009320889082219865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,1024,0.016528000434239704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,3072,0.0127697777416971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,512,0.0028817777832349143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,7168,0.036248889234330915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,768,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,2560,0.016213332613309223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,256,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,512,0.008637333081828224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,512,0.0075093333919843035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,768,0.014726221561431885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,512,0.012423111332787408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,256,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,256,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,64,0.0021253333737452826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,256,0.011000889043013254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,1536,0.010259555445777046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,128,0.00810222244924969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,32,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,128,0.0054560001525614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,65536,0.08523911237716675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,64,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,16384,0.027779555983013574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,32,0.005418666534953647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,12288,0.01851466629240248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,65536,0.3019057909647624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,10240,0.014911111858155994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,51200,0.23471999168395996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,16384,0.07841600312127008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,8192,0.012247111234400006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,768,0.008356444537639618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,12288,0.05965244770050049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,7168,0.010987555815113915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,6144,0.009353777600659264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,5120,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,10240,0.05053244365586174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,128,0.002201777779393726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,8192,0.041181332535213895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,4096,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,7168,0.03623022304640876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,3584,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,6144,0.031390223238203264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,3072,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,2560,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,5120,0.027049778236283198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,2048,0.004509333521127701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,1536,0.00388711111413108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,4096,0.022980444961124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,3584,0.02055911057525211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,1024,0.0032071111102898917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,3072,0.018242667118708294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,768,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,2560,0.016198222835858662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,51200,0.07010400295257568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,2048,0.014157333307796054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,256,0.002481777841846148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,1536,0.011959111524952782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,128,0.008383110993438298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,128,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,64,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,1024,0.009639110830095079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,768,0.008319999608728621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,512,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,256,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,128,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,64,0.0052915554907586836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,32,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,51200,0.06944088803397284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,16384,0.025361778007613286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,12288,0.01759022143152025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,65536,0.2988079918755425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,51200,0.2348568969302707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,10240,0.011414222419261932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,12288,0.059656891557905406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,10240,0.05039733317163256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,7168,0.009090666969617208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,8192,0.04085866610209147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,6144,0.008271111382378472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,32,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,5120,0.011219555305110084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,65536,0.08674399720297919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,7168,0.03623022304640876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,4096,0.009399111072222391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,6144,0.031531555785073176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,3584,0.008623999853928884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,3072,0.00776711106300354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,5120,0.02685244381427765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,2560,0.006824888702895906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,2048,0.004534222185611725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,4096,0.022582223018010456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,3584,0.021328000558747187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,1536,0.0038622220357259116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,16384,0.07830222447713216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,3072,0.01904800037542979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,1024,0.0032177778581778207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,2560,0.016521778371598985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,768,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,512,0.002675555646419525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,8192,0.009841777384281158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,256,0.0023351110931899813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,128,0.0022071111533376905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,2048,0.014136888914638095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,64,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,1536,0.012098666694429187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,1024,0.009463110731707679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,768,0.008372444245550368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,512,0.00701155596309238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,512,0.002542222125662698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,256,0.005828444328573015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,128,0.005396444350481033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,64,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,32,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,32,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,16384,1.6915831036037867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,12288,1.2150746451483834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,12288,3.6679954528808594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,16384,5.185816023084852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,12288,3.829676310221354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,16384,5.658547719319661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,10240,1.0256168577406142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,10240,3.209716373019748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,51200,6.442579481336806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,10240,3.468334197998047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,8192,2.6006452772352433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,8192,0.86117156346639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,8192,2.495663113064236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,7168,0.6957955360412598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,6144,0.6155341996086968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,7168,2.274010764227973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,6144,1.8413297865125868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,6144,1.8883715735541449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,7168,2.2037039862738714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,5120,1.6552249060736761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,5120,0.5354542202419704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,4096,0.43409066730075413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,4096,1.2382390764024522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,3584,0.37793244255913633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,5120,1.5964879989624023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,4096,1.1983199649386935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,3584,1.104397349887424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,3584,1.1146222220526802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,3072,0.9375528759426541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,2560,0.8140062226189507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,2560,0.29750490188598633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,3072,0.9467599656846789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,2560,0.7913235558403863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,2048,0.620614210764567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,2048,0.2533377806345622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,1536,0.48064708709716797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,3072,0.34095199902852374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,1024,0.3216328885820177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,2048,0.7336106830173068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,1536,0.21283111307356092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,1536,0.49741331736246747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,1024,0.15922489431169298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,768,0.28413867950439453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,512,0.17904088232252333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,768,0.14599377579159206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,1024,0.4021448824140761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,256,0.1411155594719781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,768,0.30697867605421275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,512,0.15817422337002224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,128,0.12363110648261176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,256,0.11145867241753472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,512,0.2504471143086751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,256,0.22961155573527017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,128,0.14736266930898032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,64,0.09546489185757107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,128,0.2147679991192288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,32,0.10115022129482693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,64,0.21879911422729492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,32,0.22324355443318686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,51200,15.854281955295138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,16384,3.8114000956217446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,65536,6.769461313883464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,51200,4.838839213053386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,16384,1.3251671261257596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,51200,24.95227559407552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,16384,4.355497572157118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,12288,3.4241093529595266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,12288,0.9457119835747613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,51200,12.292025248209635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,10240,0.7855360243055555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,10240,2.6675342983669705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,12288,3.275728013780382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,8192,0.6283689075046116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,10240,2.616917292277018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,8192,2.245541254679362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,7168,0.5734977722167969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,8192,2.0568514929877386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,7168,2.116536882188585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,65536,15.670756022135416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,6144,0.49282222323947483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,7168,1.7809581756591797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,6144,1.4854710896809895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,5120,0.44799375534057617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,5120,1.3772658242119684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,6144,1.4925350613064234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,4096,0.9394782384236654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,4096,0.3397386603885227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,5120,1.209453370836046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,3584,0.3057768874698215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,4096,0.952357292175293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,3584,0.8919529385036893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,3584,0.8212515513102213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,3072,0.7702159881591797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,3072,0.2650284502241346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,2560,0.5676986906263564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,3072,0.7090729077657064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,2560,0.23901600307888457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,2048,0.544139544169108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,2048,0.20155733161502412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,1536,0.3716622193654378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,2560,0.6222240130106608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,2048,0.5016933547125922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,1024,0.2469431029425727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,1536,0.3891768985324436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,768,0.19224711259206137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,1024,0.12708000342051187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,51200,18.80973646375868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,1536,0.17943021986219618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,768,0.1129431062274509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,512,0.15462488598293728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,256,0.1109315554300944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,512,0.11230222384134929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,768,0.2503173351287842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,128,0.09357600079642402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,512,0.19725333319769967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,128,0.08817422389984131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,64,0.0837964415550232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,256,0.17825510766771105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,128,0.17341332965426973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,32,0.08029955625534058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,64,0.17250488864050972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,32,0.177184886402554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,1024,0.3294666608174642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,256,0.09153599871529473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,16384,1.2082000308566623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,65536,1.9414257473415797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,51200,1.6426017549302843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,16384,0.5823724534776475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,65536,23.545778062608505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,16384,1.4004969067043727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,12288,1.145690706041124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,12288,0.30644623438517254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,51200,4.238143073187934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,10240,0.25721777809990776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,12288,1.0445280075073242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,10240,0.824127091301812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,65536,5.039541456434462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,10240,0.8552195760938855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,8192,0.21021689309014213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,8192,0.7068799866570367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,7168,0.5942231284247504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,7168,0.18861156039767793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,8192,0.6645938025580512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,6144,0.5312533378601074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,6144,0.17566488848792183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,7168,0.5877991252475315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,5120,0.3877386781904433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,5120,0.17383821805318198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,4096,0.29992977778116864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,5120,0.4044346544477675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,4096,0.12483733230166966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,3584,0.34186310238308376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,4096,0.32459823290507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,3584,0.10370666450924343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,3584,0.31146489249335396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,3072,0.26866311497158474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,3072,0.09182311428917779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,2560,0.2010951174630059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,3072,0.23709421687655977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,2560,0.08057955238554212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,2048,0.15236622757381862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,51200,7.195797390407986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,2560,0.25882310337490505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,2048,0.06845421923531426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,1536,0.119759996732076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,1024,0.09086133374108209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,2048,0.1667333311504788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,1536,0.056998219754960805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,768,0.07163288858201769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,1536,0.14578844441307917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,1024,0.04691288868586222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,1024,0.1078408890300327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,512,0.05139466789033678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,768,0.04292444388071696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,256,0.03516355488035414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,768,0.08346399996015762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,512,0.04073066512743632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,128,0.032136887311935425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,256,0.03639644384384155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,512,0.0684666633605957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,64,0.02629511058330536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,256,0.06164089176389906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,128,0.034937779108683266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,128,0.05918044514126248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,64,0.05940622091293335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,32,0.05993777513504028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,6144,0.6398631201850044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,65536,7.788465711805556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,32,0.02801688843303257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,16384,0.9036284552680122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,51200,1.1308968861897786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,65536,1.5252595477634008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,16384,1.0623075697157118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,16384,0.3112328847249349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,12288,0.23855111334058973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,12288,1.021625730726454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,12288,0.929991086324056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,10240,0.6951173146565756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,10240,0.22977156109280059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,51200,3.21832275390625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,65536,3.719933403862847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,8192,0.1626782284842597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,10240,0.6260026825798882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,8192,0.5104506810506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,7168,0.47171910603841144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,7168,0.14480000072055393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,8192,0.6095235612657335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,6144,0.37341422504848903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,6144,0.1464746660656399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,5120,0.3144071102142334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,7168,0.4403724405500624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,6144,0.41022488805982804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,4096,0.2564977804819743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,5120,0.3084746731652154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,4096,0.08958488702774048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,3584,0.19894577397240532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,3072,0.17715111043718126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,4096,0.27029866642422146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,3584,0.08124711116154988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,51200,4.6068878173828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,2560,0.16408177216847739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,3072,0.07383200195100573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,3584,0.234590212504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,2560,0.07223200135760836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,2048,0.1253395610385471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,2560,0.157441775004069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,1536,0.09849599997202556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,2048,0.05471466647254097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,1024,0.06841955582300822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,2048,0.13488799995846218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,768,0.048603554566701256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,1536,0.10223911205927531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,1024,0.04000799854596456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,1024,0.07930844359927706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,512,0.037500444385740496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,65536,5.870532565646702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,768,0.03512710995144314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,256,0.03153689040078057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,768,0.06292711363898383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,512,0.032820443312327065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,128,0.023631999890009563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,512,0.0532257788711124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,256,0.03272977802488539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,256,0.04786666565471225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,64,0.019919999771647982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,128,0.028913779391182795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,32,0.02032977839310964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,128,0.04615911179118686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,3072,0.18851289484235975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,64,0.04619910981920031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,32,0.04623022344377306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,1536,0.044682668315039746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,5120,0.1156960063510471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,16384,0.8008400069342719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,65536,1.2622168858846028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,51200,0.9677893320719401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,16384,0.2953706582387288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,12288,0.20116178194681802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,51200,2.418444527520074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,12288,0.7253697713216146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,16384,0.9638257556491427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,12288,0.6264782481723362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,10240,0.21558133761088052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,10240,0.5531315803527832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,8192,0.42766488922966855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,10240,0.5407004356384277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,8192,0.4101697868771023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,7168,0.44931379954020184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,7168,0.12366399500105117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,6144,0.26726312107510036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,7168,0.399399995803833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,6144,0.10792089170879787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,5120,0.22484978040059408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,6144,0.3401031229231093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,51200,3.978038363986545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,4096,0.18168089124891493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,5120,0.26179554727342397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,65536,3.229486253526476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,4096,0.08619378010431926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,3584,0.16776622666252983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,8192,0.13761688603295222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,4096,0.2067777713139852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,3072,0.14720443884531656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,3584,0.06918755504820082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,3072,0.06196000178654989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,2560,0.1265982256995307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,3584,0.17870844735039604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,2048,0.10209777620103623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,3072,0.1626017755932278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,1536,0.07566489113701715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,2048,0.04663911130693224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,2560,0.13834933439890543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,1024,0.05472977624999153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,2048,0.11150577333238389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,1536,0.03903466794225905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,768,0.04389155573315091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,1536,0.0875999993748135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,1024,0.03302844365437826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,512,0.030749334229363337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,5120,0.09252088599734837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,768,0.030970666143629286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,256,0.022227555513381958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,768,0.05408089028464424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,65536,5.954348246256511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,128,0.019296889503796894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,256,0.025568000144428674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,64,0.01686755485004849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,256,0.040582223070992365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,128,0.0247688889503479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,32,0.017886221408843994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,128,0.0477422210905287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,64,0.039157334301206805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,2560,0.05424355467160543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,32,0.03951111104753282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,1024,0.06885244448979695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,512,0.02808444367514716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,512,0.045422222879197865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,16384,0.5455750889248318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,65536,1.1388222376505535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,51200,0.8660817676120334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,16384,0.707150247361925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,16384,0.23994665675693086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,12288,0.17681511243184408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,12288,0.6969413227505155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,51200,2.229649861653646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,10240,0.3963644504547119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,10240,0.19121778011322021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,8192,0.31571732627020943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,65536,2.9039590623643665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,12288,0.8203804228040906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,10240,0.4235511091020372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,8192,0.37447555859883624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,8192,0.14466133382585314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,7168,0.2895955509609646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,6144,0.24594399664137098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,5120,0.20562667316860625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,6144,0.0909608867433336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,7168,0.3147795465257433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,51200,3.1694488525390625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,6144,0.25209776560465497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,5120,0.08851466576258342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,4096,0.07280977567036946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,3584,0.13319733407762316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,5120,0.2757013373904758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,3072,0.11399022738138835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,3584,0.06148711178037855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,3584,0.1542266607284546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,2560,0.09909066889021133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,3072,0.05238133337762621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,3072,0.13514310783810085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,2048,0.0834524432818095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,2560,0.04728088776270548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,65536,4.333261277940538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,2560,0.11046933465533787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,1536,0.07097599903742473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,2048,0.03972711165746053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,7168,0.10645333263609146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,2048,0.09294133053885566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,1024,0.04524622360865275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,768,0.03554755449295044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,1536,0.07238044341405232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,4096,0.15074933899773493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,1024,0.05431822273466322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,512,0.02444177700413598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,768,0.02493155499299367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,768,0.044806222120920815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,256,0.01755199995305803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,512,0.022612444228596155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,4096,0.16830489370557997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,128,0.016004444824324716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,512,0.035918222533331975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,256,0.020972443951500788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,64,0.013718222578366598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,256,0.03158222304450141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,32,0.014456000592973499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,128,0.030227555168999568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,64,0.03018222252527873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,32,0.029832889636357624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,1536,0.038224889172448054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,1024,0.028424001402325098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,128,0.020071110791630216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,65536,0.9848933749728732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,51200,0.7540417777167426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,16384,0.5613999896579319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,16384,0.27109511693318683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,51200,1.7721182505289714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,16384,0.6209262212117513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,12288,0.49549955791897243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,12288,0.21267910798390707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,10240,0.38161688380771214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,65536,2.523359934488932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,10240,0.1679617828792996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,12288,0.552353752983941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,8192,0.3123573462168376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,10240,0.3759377797444661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,8192,0.11421244674258763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,7168,0.2885519928402371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,7168,0.09399466382132636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,6144,0.2109004391564263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,7168,0.26605243153042263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,8192,0.358841790093316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,6144,0.1075635552406311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,4096,0.15202400419447157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,5120,0.07191733519236247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,51200,2.9948736826578775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,3584,0.12426577674018012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,5120,0.19440355565812853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,4096,0.06122222211625841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,65536,3.4364941914876304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,4096,0.17348266972435844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,3072,0.10236089097128974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,3584,0.05316977699597677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,2560,0.0912933349609375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,3072,0.05419111251831055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,2048,0.07666044765048556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,2560,0.04123377799987793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,1536,0.05718311336305407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,2048,0.03570400012864007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,2048,0.08080089092254639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,1024,0.03848355677392747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,1536,0.03001333276430766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,1536,0.06445955567889743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,5120,0.2000497712029351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,768,0.03355377912521362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,1024,0.025335111551814612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,6144,0.23116800520155165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,1024,0.04920088913705614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,512,0.023231999741660223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,256,0.016591111818949383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,768,0.04013955593109131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,512,0.020267556111017864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,128,0.013472889032628802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,512,0.0321413344807095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,256,0.01828977796766493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,256,0.028183999988767836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,64,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,3584,0.1436177757051256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,128,0.018767111831241183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,3072,0.1151768896314833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,128,0.026913776993751526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,64,0.02676888969209459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,32,0.02664088871743944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,768,0.022224000758594934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,65536,0.8284622298346626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,32,0.01239733315176434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,2560,0.09792533185746934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,51200,1.3826399909125433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,51200,0.6551164521111382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,16384,0.5203368928697374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,16384,0.18407377931806776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,12288,0.38720088534884983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,16384,0.5373742315504286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,12288,0.16462133990393743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,65536,2.2527217864990234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,12288,0.44423643747965497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,10240,0.34543289078606504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,10240,0.11407555474175347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,8192,0.2946995629204644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,10240,0.33256710900200737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,8192,0.10753333568572998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,7168,0.22968443234761557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,51200,2.5505449506971574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,7168,0.08228000005086263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,8192,0.2623671160803901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,6144,0.19881778293185762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,65536,2.90704345703125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,5120,0.1528408924738566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,4096,0.11565689245859782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,7168,0.3088151084052192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,5120,0.06884355677498712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,6144,0.22511466344197592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,3584,0.10627822081247966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,4096,0.05689866675270928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,5120,0.16430755456288657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,3072,0.09241777658462524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,3584,0.04875911275545756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,2560,0.07664444711473253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,3072,0.04134755664401584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,3584,0.12044355604383682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,2048,0.06146133608288235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,3072,0.10252444611655341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,2560,0.03829333186149597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,1536,0.04925600025388929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,2560,0.08828177716996934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,2048,0.030809776650534734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,1024,0.032476445039113365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,2048,0.07219466898176405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,768,0.026407111022207472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,1536,0.026643556025293138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,1024,0.021927111678653296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,512,0.01940266622437371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,768,0.01958933307064904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,1024,0.04274844461017185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,256,0.014122666584120857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,6144,0.08450844552781846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,768,0.036143110858069524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,128,0.012433778080675336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,512,0.017835555805100333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,512,0.028565333949195013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,64,0.011976888610257043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,256,0.024661334024535284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,128,0.01648000048266517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,4096,0.1303102175394694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,128,0.023696889479955036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,64,0.02365066607793172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,32,0.023634667197863262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,1536,0.057872891426086426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,65536,0.7028809123569064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,256,0.016528000434239704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,32,0.012553777959611682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,51200,1.1636408699883354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,51200,0.5559288660685221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,16384,0.444910208384196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,65536,1.5648132960001628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,16384,0.15680266751183405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,12288,0.300872008005778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,16384,0.4498444663153754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,12288,0.14871644973754883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,10240,0.3032240072886149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,10240,0.11379110813140869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,12288,0.3515137831370036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,8192,0.2437831030951606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,10240,0.2707173294491238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,8192,0.09118666913774277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,7168,0.20206667317284477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,8192,0.23936266369289824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,51200,2.0309990776909723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,5120,0.1306355529361301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,6144,0.06704799996482001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,7168,0.19390400250752768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,6144,0.1858373350567288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,4096,0.10364800029330784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,5120,0.05405066741837395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,65536,2.785750283135308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,5120,0.15053510665893555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,3584,0.08697333600785996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,4096,0.048787554105122886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,3072,0.08179110950893827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,4096,0.11151733663347031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,3584,0.04569333460595873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,2560,0.06378400325775146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,3072,0.03533600105179681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,2048,0.05150666501786974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,2560,0.036348445547951594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,3072,0.08717599842283462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,1536,0.04251199960708618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,2560,0.07361600134107801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,1024,0.029554665088653564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,2048,0.0623555580774943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,1536,0.023423999547958374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,1536,0.04907110995716519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,6144,0.1626195510228475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,1024,0.01924533314175076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,1024,0.03702488872740004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,768,0.01761866609255473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,512,0.01647111111217075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,768,0.030697776211632624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,512,0.01610488858487871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,256,0.01236088905069563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,512,0.024883555041419134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,128,0.009472000102202097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,256,0.015115555789735583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,256,0.021382222572962444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,3584,0.10053866439395481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,128,0.01496355566713545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,64,0.008622222476535374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,32,0.008656000097592672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,128,0.02048711147573259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,2048,0.027447111076778833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,64,0.020611556039916146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,32,0.020474667350451153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,7168,0.07281866338517931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,65536,0.5641884273952907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,51200,0.4430053234100342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,51200,0.8871039814419217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,16384,0.36379467116461855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,65536,1.3262497584025066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,16384,0.13992622163560656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,12288,0.2683662308586968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,768,0.021684444612926904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,16384,0.42494932810465497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,12288,0.10040444135665894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,10240,0.19371378421783447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,51200,1.4178560045030382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,12288,0.3569600052303738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,8192,0.15510577625698513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,10240,0.08438044124179417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,7168,0.13697689109378391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,65536,1.9872826470269098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,10240,0.3100071218278673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,6144,0.12657333744896784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,8192,0.1851066615846422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,5120,0.10989777909384833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,6144,0.060023109118143715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,7168,0.18060178226894805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,4096,0.0833466649055481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,5120,0.04576889011594984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,6144,0.15017599529690212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,3584,0.07339111301634046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,4096,0.03806489043765598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,5120,0.11625866095225017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,3072,0.06230400005976359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,4096,0.09932444492975871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,3584,0.03366133239534166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,3584,0.08342844247817993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,3072,0.030488000975714788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,2048,0.0406328903304206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,2560,0.027642667293548584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,3072,0.07297511233223809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,1536,0.0336168905099233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,2560,0.0631200008922153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,2048,0.02347822156217363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,1024,0.023722666833135817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,2048,0.05203466614087423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,1536,0.019983111156357657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,768,0.019036443697081674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,1536,0.04129066732194688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,8192,0.06818666723039415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,512,0.013757333159446716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,768,0.01570577753914727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,7168,0.06178310844633314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,256,0.010664000279373594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,768,0.0262444449795617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,512,0.014384888940387301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,512,0.020984888076782227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,128,0.008879111044936711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,64,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,256,0.018215111560291715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,128,0.01346577786737018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,32,0.008054222497675154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,128,0.017529777354664274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,2560,0.05417955583996243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,64,0.01751733322938283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,32,0.017500445246696472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,1024,0.016844444804721408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,65536,0.497378667195638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,1024,0.031341334184010826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,256,0.013744889034165276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,51200,0.9193252987331815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,65536,1.2524195777045357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,51200,0.3939351240793864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,16384,0.11566844251420762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,12288,0.23172444767422148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,12288,0.0907848874727885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,10240,0.17666221989525688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,51200,1.2423057556152344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,8192,0.12492088476816814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,10240,0.0783440007103814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,12288,0.3500666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,10240,0.27656621403164333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,7168,0.14151199658711752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,8192,0.06328088707394071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,8192,0.17042844825320771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,6144,0.11108444134394328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,7168,0.05577066871854994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,7168,0.14830399884117976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,5120,0.09884978002972072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,6144,0.04923999971813626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,16384,0.2857288784450955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,4096,0.07403110795550875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,6144,0.1273039976755778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,5120,0.04352444410324097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,5120,0.10809155305226643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,3584,0.06954844130410089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,4096,0.034344888395733304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,3072,0.05103111267089844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,4096,0.08490577671262954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,2560,0.040227555566363864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,3584,0.07753866910934448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,3072,0.028024888700909082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,65536,1.6560871336195204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,3072,0.06549955738915338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,2560,0.025748444928063288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,2560,0.05628444751103719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,2048,0.021709332863489788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,2048,0.047127111090554126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,1024,0.01925155520439148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,1536,0.018479110466109384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,1536,0.03760533200369941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,768,0.016141333513789706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,1024,0.015863110621770222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,1024,0.02829244401719835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,512,0.012155555188655853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,768,0.014833778142929077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,16384,0.32407821549309623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,768,0.023552889625231426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,256,0.009008888569143083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,512,0.013767999907334646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,3584,0.03092799915207757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,256,0.012904889053768583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,128,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,64,0.006908444480763541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,256,0.016499555773205228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,2048,0.035312000248167254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,128,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,32,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,128,0.016040000650617812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,1536,0.029574222034878198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,64,0.015698666373888653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,32,0.016199111938476562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,65536,0.4366746743520101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,51200,0.3425750997331407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,16384,0.25108978483412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,51200,0.6475955645243326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,65536,0.8203528722127279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,12288,0.1875715520646837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,512,0.018907555275493197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,12288,0.089629332224528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,10240,0.17123999860551622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,16384,0.3285493320888943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,12288,0.22771822081671822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,10240,0.07427199681599934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,51200,1.0525617599487305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,8192,0.1374151176876492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,65536,1.409189330206977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,8192,0.05576088693406847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,10240,0.17774221632215711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,7168,0.12470222844017877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,6144,0.09847911198933919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,5120,0.08070755667156644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,8192,0.14630399809943304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,6144,0.042746666404936046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,4096,0.061662223603990346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,7168,0.13670489523145887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,5120,0.037647111548317805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,6144,0.11118133862813313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,3584,0.056553781032562256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,4096,0.030795554320017498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,5120,0.0974844429228041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,3072,0.045884443653954395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,4096,0.07521422041787042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,3584,0.028304000695546467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,3584,0.06633866495556302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,3072,0.026166222161716882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,2048,0.032307555278142296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,2560,0.023073777556419373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,16384,0.10467111402087742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,3072,0.05770311090681288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,1536,0.025847110483381484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,1024,0.018783110711309645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,2560,0.04955733484692044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,2048,0.020117332537968952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,1536,0.017663111289342243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,768,0.015216888652907478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,2048,0.0424213343196445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,1024,0.015239111251301236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,512,0.011426666544543372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,768,0.014159111513031853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,1024,0.025389333566029865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,256,0.009710222482681274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,512,0.013122667041089801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,7168,0.056482666068606906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,512,0.01720711092154185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,128,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,256,0.012379555238617791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,256,0.014818666709793938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,64,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,32,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,128,0.014667555689811707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,2560,0.041618665059407554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,64,0.014153778553009033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,32,0.01443555619981554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,1536,0.033687111404207014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,65536,0.38591644499037003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,768,0.021351110604074266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,51200,0.5858506626553005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,65536,0.8569395807054309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,16384,0.2203813393910726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,128,0.01239466667175293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,51200,0.3128142091963026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,65536,1.1074480480617948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,12288,0.1807911131117079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,16384,0.264972448348999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,12288,0.0820853312810262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,10240,0.145306666692098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,8192,0.11502133475409614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,12288,0.183716442849901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,51200,0.8534880214267306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,10240,0.16962756050957573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,7168,0.10144622458351983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,8192,0.05199911197026571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,7168,0.04360533422893948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,6144,0.0802639987733629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,8192,0.13599199718899196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,5120,0.06523822413550483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,6144,0.03818755679660373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,7168,0.1166728867424859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,4096,0.04979644550217522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,5120,0.034726222356160484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,6144,0.09537155760659112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,3584,0.04296711087226868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,4096,0.028051555156707764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,5120,0.08357244729995728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,3584,0.025115556187099878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,4096,0.0670071111785041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,16384,0.10972177982330322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,2560,0.031911111540264554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,3072,0.02278577784697215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,3584,0.0589546693695916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,2048,0.026087111896938745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,1536,0.022114665971861944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,3072,0.05134755704138014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,2560,0.020395555429988436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,2048,0.018256000346607633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,1024,0.016041777200169034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,1536,0.01587377819750044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,2048,0.03735555542839898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,10240,0.06454044580459595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,768,0.01312355531586541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,1024,0.013632000320487551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,512,0.010106667048401302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,1024,0.022495110829671223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,256,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,768,0.018992000155978732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,512,0.012208888928095499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,128,0.006353777729802662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,512,0.01551555593808492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,256,0.011398221883508893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,64,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,128,0.011398221883508893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,3072,0.03973955578274197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,32,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,128,0.012789333032237159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,64,0.012713777522246042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,32,0.0127697777416971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,2560,0.044525331921047635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,65536,0.6111297607421875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,65536,0.3186364438798693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,51200,0.4514266649881999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,768,0.013236444857385425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,16384,0.15782666206359863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,12288,0.12371910942925347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,51200,0.26704443825615776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,65536,0.8917840321858724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,256,0.01352533366945055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,10240,0.106331553724077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,16384,0.22749956448872885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,12288,0.06875822279188368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,51200,0.7014035648769803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,8192,0.09023021989398533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,10240,0.05707999732759264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,12288,0.17341689268747965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,7168,0.07995822032292683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,10240,0.1342302163441976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,8192,0.04298311140802172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,6144,0.06876444154315524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,7168,0.04046133491728041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,5120,0.05677511294682821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,8192,0.11051199833552043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,6144,0.03271377748913235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,4096,0.043953776359558105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,7168,0.0975582202275594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,5120,0.02912799848450555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,6144,0.08222222328186035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,1536,0.0296106669637892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,3584,0.0377200014061398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,4096,0.024688000480333965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,3072,0.033245331711239286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,4096,0.05652533637152778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,2560,0.028661333852344092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,3072,0.020541333489947848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,3584,0.0505582226647271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,16384,0.07928888665305243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,3072,0.04519644379615784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,2560,0.018576888574494254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,1536,0.018833777970737882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,2560,0.03902133305867513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,2048,0.016748444901572335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,1024,0.013447999954223633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,2048,0.03285333183076646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,1536,0.014835556348164877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,1536,0.0262746661901474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,768,0.011160000330872007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,1024,0.012967111335860359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,512,0.008850666383902231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,1024,0.019955555597941082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,768,0.012081777883900536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,256,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,768,0.01664799948533376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,512,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,128,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,512,0.013829333086808523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,5120,0.07088088989257812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,256,0.0107004443804423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,64,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,3584,0.0221742226017846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,32,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,128,0.010748444332016839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,128,0.011429333024554782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,2048,0.02365155518054962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,64,0.011048888994587792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,32,0.011384000380833944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,65536,0.5179057651095921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,51200,0.3815440071953668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,51200,0.2086248927646213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,16384,0.13467821809980604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,12288,0.1040266686015659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,51200,0.5474444495307075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,65536,0.6956017812093099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,10240,0.07632444302241008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,16384,0.19235377841525605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,12288,0.07291377915276422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,8192,0.0646631121635437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,256,0.011945777469211154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,12288,0.13054133786095515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,10240,0.05071022113164266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,7168,0.05638933181762695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,10240,0.1095937755372789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,8192,0.03849777910444472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,6144,0.050203555160098605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,7168,0.03364177876048618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,5120,0.04158577654096816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,6144,0.029351999362309773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,7168,0.07809511158201429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,4096,0.03233333428700765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,6144,0.06788088877995808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,3584,0.02823644545343187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,65536,0.2591128879123264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,4096,0.02139288849300808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,3072,0.025115556187099878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,4096,0.04655288987689548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,3584,0.01910044418440925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,3584,0.04169155491722954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,2560,0.02141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,3072,0.017677333619859483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,16384,0.06891022125879924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,3072,0.03686044282383389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,2560,0.016746666696336534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,1536,0.013796444568369122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,2560,0.031854222218195595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,2048,0.01569955547650655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,1024,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,2048,0.026672000686327618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,1536,0.013251555462678274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,768,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,1536,0.02205955485502879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,1024,0.011600889265537262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,8192,0.11426666047837998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,512,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,1024,0.016886222693655226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,768,0.011054221954610614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,5120,0.02548622257179684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,512,0.010326222413116032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,768,0.014674666855070325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,128,0.004591111093759537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,512,0.012040889097584618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,256,0.00980355590581894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,64,0.004239999999602635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,256,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,128,0.009888889061080085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,128,0.01001422190003925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,2048,0.018165333403481376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,64,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,32,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,65536,0.33538132243686253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,65536,0.2060328854454888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,51200,0.28902578353881836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,16384,0.09341777695549859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,256,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,51200,0.1647084421581692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,65536,0.542409790886773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,12288,0.06996088557773165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,5120,0.057258665561676025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,51200,0.4490470886230469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,16384,0.05522933271196154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,16384,0.14845155345069036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,10240,0.058506667613983154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,32,0.004579555657174853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,12288,0.044284443060557045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,12288,0.10485955741670396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,10240,0.03870933254559835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,7168,0.0407955547173818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,8192,0.028295109669367473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,6144,0.03515555461247762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,10240,0.0892817775408427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,7168,0.02534399926662445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,5120,0.027635556128289964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,6144,0.022335999541812476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,7168,0.06269777483410306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,5120,0.020132443971104093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,6144,0.05443111062049866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,3584,0.01923288901646932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,5120,0.046126223272747464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,4096,0.017364444004164804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,3072,0.016759110821617972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,4096,0.03813066747453477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,3584,0.015878222054905362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,2560,0.014491554763582019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,3584,0.03426311082310147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,3072,0.01481599940194024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,2048,0.012492444780137805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,3072,0.0306000014146169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,1536,0.009988444546858469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,2560,0.013757333159446716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,2048,0.012416889270146688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,2560,0.026203556193245783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,1024,0.007919111185603673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,2048,0.02236266599761115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,768,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,1024,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,1536,0.018479110466109384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,512,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,1024,0.014093332820468478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,8192,0.0712604456477695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,768,0.009983110758993361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,256,0.004546666724814309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,768,0.012423999607563019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,4096,0.021664000219768945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,128,0.0038311113086011675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,256,0.00903200027015474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,512,0.010628444453080496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,64,0.0035333335399627686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,128,0.009005332986513773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,32,0.003568888952334722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,128,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,64,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,65536,0.2507431242201063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,32,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,51200,0.19388888941870797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,8192,0.04795644349522061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,65536,0.20037244425879586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,1536,0.011379555695586734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,65536,0.4564817746480306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,51200,0.178001774681939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,12288,0.05223200056287977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,16384,0.054767999384138316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,16384,0.11939644813537598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,10240,0.044807109567854136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,512,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,12288,0.04369422131114536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,8192,0.03410666518741184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,12288,0.09381066428290473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,10240,0.03827822208404541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,256,0.008723555339707268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,7168,0.028743111424975928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,10240,0.07545244693756104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,6144,0.02572622232966953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,8192,0.06070578098297119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,7168,0.0251875552866194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,5120,0.021747555997636583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,6144,0.02256444427702162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,7168,0.05368444323539734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,4096,0.01793155570824941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,5120,0.019650666250122916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,6144,0.047149333688947886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,3584,0.015592889653311836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,5120,0.039956443839603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,4096,0.01720888912677765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,3072,0.013735999663670858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,16384,0.0724088880750868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,4096,0.032962666617499456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,2560,0.012075555821259817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,3584,0.02955555584695604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,3072,0.014645333091417948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,2048,0.010262221925788457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,51200,0.3620737658606635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,3072,0.026698667142126296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,1536,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,2560,0.013753777576817406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,2048,0.012418666647540199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,1024,0.00654755574133661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,2048,0.019573332534896005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,1536,0.011431111229790581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,1536,0.016176889340082806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,768,0.005809777726729711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,1024,0.01034222212102678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,8192,0.027662222584088642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,512,0.004671999977694618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,768,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,1024,0.012718222207493253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,256,0.004176889028814104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,768,0.010776888993051318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,512,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,128,0.003548444559176763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,512,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,256,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,64,0.003173333282272021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,32,0.0034977777136696708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,128,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,128,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,3584,0.0158924443854226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,64,0.007342221836249034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,65536,0.19536799854702422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,32,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,51200,0.15585066212548151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,65536,0.1539377768834432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,16384,0.05298755566279093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,2560,0.022670222653283015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,51200,0.12367110782199436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,65536,0.4019724527994792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,16384,0.043324444029066295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,51200,0.3245573308732775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,10240,0.03164088726043701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,16384,0.10543821917639838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,8192,0.024417777856190998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,12288,0.034838222795062594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,12288,0.08015910784403484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,10240,0.03122577733463711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,7168,0.02074044446150462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,6144,0.018498667412334018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,8192,0.02164177762137519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,10240,0.06815822256935967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,5120,0.016343999240133498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,256,0.007979555262459649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,7168,0.019000889526473153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,8192,0.05418311225043403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,4096,0.013034666577974955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,7168,0.04884711239073011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,5120,0.015266666809717814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,3584,0.011953777737087674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,4096,0.013439999686347114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,5120,0.03608444333076477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,3072,0.010658666491508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,4096,0.029501332177056208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,3584,0.0127368892232577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,2560,0.009383999639087254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,3584,0.026761778526835974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,3072,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,2048,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,3072,0.023817777633666992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,1536,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,2560,0.020888888173633154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,2048,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,1024,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,2048,0.017847999930381775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,1536,0.010052444206343757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,768,0.0047315553658538395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,1536,0.014767110347747803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,1024,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,512,0.003947555604908201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,1024,0.011687999798191918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,768,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,256,0.0031999999450312722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,768,0.010112889111042023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,6144,0.017112000121010672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,512,0.008611555728647444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,128,0.002806222273243798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,6144,0.04240355557865567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,256,0.008278222547637092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,64,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,256,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,128,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,128,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,65536,0.12921778361002603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,64,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,32,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,51200,0.09416444434059991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,65536,0.15329333146413168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,16384,0.03787111242612203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,65536,0.3273351192474365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,51200,0.12275821632809109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,51200,0.25579110781351727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,12288,0.027494221925735474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,16384,0.043902221653196544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,512,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,10240,0.021190222766664293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,12288,0.03490399983194139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,12288,0.06506399975882636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,32,0.0028364445186323593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,8192,0.01700888905260298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,10240,0.030092444684770372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,7168,0.015563555889659457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,8192,0.02127200033929613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,12288,0.041053334871927895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,8192,0.044385777579413525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,6144,0.01348000019788742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,7168,0.01888533267709944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,7168,0.039581331941816546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,5120,0.011626666618718041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,6144,0.017454221844673157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,4096,0.008925333619117737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,5120,0.015498666299713982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,6144,0.034635557068718806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,3584,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,5120,0.02937244375546773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,4096,0.013669333524174161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,3072,0.008617777791288164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,4096,0.02439022229777442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,16384,0.08602311213811238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,2560,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,3072,0.012294222083356647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,3072,0.01976799964904785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,2048,0.006016888966162999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,2560,0.017312889297803242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,10240,0.05452088846100701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,1536,0.005199999858935674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,2048,0.010852444503042432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,2560,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,2048,0.014866666661368476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,1024,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,1536,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,1536,0.012436444560686747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,768,0.0042088888585567474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,512,0.003485333174467087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,1024,0.010043555663691627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,768,0.008954666554927826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,256,0.0029120000286234748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,768,0.008960000342792934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,512,0.008398222426573435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,512,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,128,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,256,0.008096888661384583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,3584,0.012912888493802814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,64,0.0023786665664778813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,3584,0.022038223014937505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,256,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,32,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,128,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,128,0.005712000032265981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,65536,0.09200621975792779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,64,0.005628444254398346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,51200,0.07369422250323825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,32,0.00572622236278322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,2560,0.011728000309732227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,65536,0.15392622682783338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,12288,0.022893332772784766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,65536,0.28663733270433217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,16384,0.04487199915779961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,51200,0.12484888235727946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,10240,0.01753866672515869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,51200,0.22277421421474883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,16384,0.07487999730639987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,8192,0.013070222404268054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,12288,0.03540444374084473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,1024,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,7168,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,10240,0.03133422136306763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,8192,0.023014222582181294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,6144,0.010411555568377176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,10240,0.04823377728462219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,5120,0.009032888544930352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,7168,0.020282665888468426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,8192,0.039127998881869845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,4096,0.007668444679843054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,6144,0.018398221996095445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,7168,0.03462666604253981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,5120,0.016306666864289176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,6144,0.030216889248953924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,4096,0.014522666732470194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,5120,0.026032888227038916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,3072,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,4096,0.02195466723706987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,3584,0.01311911145846049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,2560,0.0058044443527857465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,3584,0.019666666785875957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,2048,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,16384,0.028891556792789038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,3072,0.017850667238235474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,1536,0.004273777620659934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,2560,0.01569333341386583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,2048,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,2048,0.01367200000418557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,1024,0.003536000019974179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,1536,0.010055999788973067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,1536,0.011414222419261932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,12288,0.05702310800552368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,768,0.003170666595300039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,1024,0.009338666995366415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,512,0.002814222127199173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,768,0.008971555365456475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,768,0.008340444829728868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,256,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,512,0.008578666382365758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,3584,0.007126222054163615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,128,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,512,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,256,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,64,0.002193777718477779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,256,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,128,0.007970666719807519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,32,0.002472888885272874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,3072,0.0124177775449223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,128,0.0053182223604785065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,2560,0.011762667033407422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,65536,0.08943466345469157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,51200,0.07139733102586535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,16384,0.02472711106141408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,64,0.005335999859703912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,32,0.005388444496525659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,10240,0.01204711116022534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,65536,0.2847822242312961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,16384,0.0743466681904263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,51200,0.2226480113135444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,8192,0.010835555692513784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,7168,0.010258666343159145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,10240,0.04772444566090902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,6144,0.009593778186374241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,1024,0.009345778160625035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,5120,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,4096,0.00701155596309238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,8192,0.038819554779264666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,7168,0.034212443563673235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,3584,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,3072,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,6144,0.02997422218322754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,5120,0.02595377796226078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,2560,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,2048,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,4096,0.02176888949341244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,1536,0.003906666818592283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,3584,0.019705777366956074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,1024,0.0031946665710873077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,3072,0.017596445149845548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,2560,0.015570667054918079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,768,0.0029235554652081597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,512,0.0027271111806233725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,1536,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,12288,0.01735111077626546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,256,0.002463999928699599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,128,0.0021582223061058256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,1024,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,64,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,32,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,768,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,512,0.007079111205206976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,65536,0.08544088734520806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,256,0.005631111148330901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,128,0.005377777748637729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,64,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,16384,0.02388800018363529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,51200,0.06912355290518866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,32,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,12288,0.015288889408111572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,10240,0.01145155562294854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,8192,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,65536,0.2835600111219618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,12288,0.056800001197391085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,7168,0.009290667043791877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,51200,0.22235200140211317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,6144,0.008662222160233391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,8192,0.03882222374280294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,5120,0.01071111112833023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,4096,0.008990222381220924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,7168,0.03432800041304694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,6144,0.02997866604063246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,3584,0.008282666405042013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,2048,0.01346577786737018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,5120,0.025777777036031086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,3072,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,2560,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,2048,0.004582222137186262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,3584,0.0199608885579639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,3072,0.018020444446139865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,1536,0.0038328886859946777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,2560,0.016190222567982145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,1024,0.0031831111345026228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,2048,0.013697778185208639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,1536,0.011355555719799466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,1024,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,512,0.0025617778301239014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,768,0.008061333662933773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,256,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,16384,0.07432978020773993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,128,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,10240,0.04792355497678121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,64,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,512,0.007061333292060428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,32,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,256,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,128,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,64,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,4096,0.02167822255028619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,12288,0.05687910980648465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,768,0.002883555574549569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,32,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,16384,1.5826746622721355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,16384,4.584802839491102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,16384,5.405943976508246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,51200,6.246636284722222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,12288,4.246218787299262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,16384,4.900903065999349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,12288,3.846973419189453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,12288,1.152807977464464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,10240,0.9331351386176215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,12288,3.5264809926350913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,10240,3.1281272040473094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,10240,3.0972381167941623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,8192,0.7705830997890897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,10240,2.909188376532661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,8192,2.847169876098633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,8192,2.4364187452528214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,8192,2.3534043629964194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,7168,0.699485354953342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,51200,14.810268825954862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,7168,2.5350301530626083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,7168,2.20851198832194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,6144,0.5962328910827637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,7168,2.0779830084906683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,6144,2.2636951870388455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,6144,1.8615919748942058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,5120,1.489488919576009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,6144,1.69122314453125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,5120,0.6288862228393555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,5120,1.5729653040568035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,4096,1.219395531548394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,4096,0.4136791229248047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,5120,1.4841636021931965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,4096,1.222614182366265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,3584,1.086668438381619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,4096,1.1515795389811199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,3584,0.3691111140780979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,3584,1.0883671442667644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,3072,0.938818613688151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,3072,0.32977244589063853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,3584,1.058153788248698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,2560,0.685456011030409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,3072,0.9327333238389758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,2560,0.2987119886610243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,3072,0.8823448816935221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,2048,0.6028550995720757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,2048,0.2420969009399414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,2560,0.8928470611572266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,2560,0.7167146470811633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,2048,0.6317884657118056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,1536,0.4658675723605686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,1536,0.19437777996063232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,1024,0.30411020914713544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,2048,0.5909653239780003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,51200,16.25750732421875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,1536,0.4974044693840875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,768,0.23964444796244302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,1024,0.17923466364542642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,1536,0.4546533160739475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,512,0.193284445338779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,1024,0.40453065766228574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,768,0.15334755844540066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,1024,0.3525671164194743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,256,0.1277502245373196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,768,0.27790045738220215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,768,0.30457067489624023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,512,0.25441333982679576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,128,0.10893955495622422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,256,0.10647466447618271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,512,0.2507226732042101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,256,0.23067821396721733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,64,0.08874577946133083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,128,0.10426221953497992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,256,0.17909510930379233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,128,0.21354133552975127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,32,0.09092977974149917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,128,0.13513155778249106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,64,0.21597511238522002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,32,0.21783467133839926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,51200,23.569529215494793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,512,0.12252177132500543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,51200,4.562591976589627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,65536,6.216979556613499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,16384,4.359090593126085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,51200,11.706025865342882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,16384,4.69264178805881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,16384,1.184343973795573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,65536,15.117296006944445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,16384,3.877699534098307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,12288,3.263307571411133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,12288,3.061153835720486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,51200,12.863296508789062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,12288,0.9210373560587565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,10240,0.7429982291327583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,12288,2.773395538330078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,51200,19.151318020290798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,10240,2.4483538733588324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,10240,2.4737440745035806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,8192,0.5875760184393989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,10240,2.299255159166124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,8192,2.4210906558566623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,8192,2.029802746242947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,7168,1.730796496073405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,8192,1.8171759711371527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,7168,0.6811919742160373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,7168,1.7167689005533855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,6144,0.4935306443108453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,6144,1.506598260667589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,7168,1.6151289410061305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,6144,1.4087262683444555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,65536,23.415279812282986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,65536,18.684551662868923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,5120,0.40532623396979434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,6144,1.377806239657932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,5120,1.3230293061998155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,4096,0.8832053078545464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,5120,1.2299315134684246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,4096,0.3275546762678358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,4096,0.96524535285102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,5120,1.1021315256754558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,3584,0.2977208826276991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,3584,0.8952479892306857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,3072,0.7229377958509656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,3584,0.9187502331203885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,3072,0.26139820946587455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,4096,0.8935599856906467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,3584,0.7563324504428439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,2560,0.6487493515014648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,2560,0.22742843627929688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,3072,0.702240890926785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,2048,0.4970729086134169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,3072,0.6607413291931152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,2048,0.1920106675889757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,1536,0.3421671125623915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,2560,0.5581964386834039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,2048,0.5129297574361166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,1536,0.15610488255818686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,2560,0.633645322587755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,1024,0.24549600813123915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,1536,0.5285369025336372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,2048,0.5244728724161783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,1024,0.1247004403008355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,768,0.20369244946373832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,1024,0.30603734652201336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,1536,0.36300889650980633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,512,0.15602310498555502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,768,0.11828800042470296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,256,0.11133689350552028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,512,0.09798755248387654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,512,0.2005386617448595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,1024,0.26623378859625924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,768,0.2211386627621121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,128,0.08182044161690606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,256,0.08551022079255845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,512,0.18003288904825845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,64,0.07110666566424899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,128,0.08337066570917766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,32,0.07235911157396105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,128,0.16676978270212808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,256,0.1457351048787435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,128,0.10679822497897679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,64,0.17039999696943495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,32,0.17120444774627686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,768,0.319376892513699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,256,0.17941421932644316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,65536,1.9160372416178386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,51200,1.4461137983534071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,51200,3.985378689236111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,16384,1.3222462336222331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,16384,1.6339279810587566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,16384,0.3959466616312663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,65536,5.147100236680773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,12288,1.0226667192247179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,16384,1.3111733330620658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,51200,4.055826822916667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,12288,0.9992693795098199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,12288,0.3078133265177409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,51200,6.155677371554905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,12288,0.8597653177049426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,10240,0.25403467814127606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,10240,0.8486888673570422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,8192,0.6396302117241753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,10240,0.8377617730034722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,65536,5.553734249538845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,10240,0.7046293152703179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,8192,0.24812889099121094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,7168,0.5984400113423666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,8192,0.5493769115871853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,65536,8.422337002224392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,7168,0.5476506551106771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,6144,0.4468400213453505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,6144,0.15853956010606554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,7168,0.6332106590270996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,6144,0.4932390848795573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,5120,0.3668817679087321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,8192,0.6633262104458278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,5120,0.1353822284274631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,6144,0.428056001663208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,7168,0.18199377589755586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,4096,0.2948959933386909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,5120,0.387135108311971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,4096,0.11168177922566731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,5120,0.3589937686920166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,3584,0.2573964330885145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,4096,0.3083057668473986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,3584,0.10076089037789239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,3072,0.22905243767632377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,4096,0.28640977541605633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,3584,0.30045776897006565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,3072,0.08882666958702935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,3072,0.23373866081237793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,3584,0.2519208855099148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,2048,0.14839911460876465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,3072,0.22315822707282174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,2560,0.2018515533871121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,1536,0.127618670463562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,2048,0.0670328868760003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,2048,0.1647999948925442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,2560,0.19012622038523355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,1024,0.07657688856124878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,1536,0.13082222143809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,2048,0.15699733628167048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,768,0.06615110900666979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,1024,0.0997848908106486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,1024,0.04587733414438036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,1536,0.13965688811408147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,512,0.048785779211256236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,768,0.042303999265034996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,1024,0.09764266676372951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,768,0.07931822538375854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,256,0.03107644451989068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,2560,0.19627643956078422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,512,0.03861600160598755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,768,0.08192000124189588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,512,0.06812177764044867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,128,0.025552888711293537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,2560,0.0783991085158454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,256,0.034062223301993474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,256,0.06092266903983223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,64,0.02311466634273529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,512,0.06793066528108385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,32,0.023895111348893907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,128,0.0339680016040802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,128,0.05791199869579739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,256,0.05287733342912462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,1536,0.055440889464484326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,128,0.04048177931043837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,64,0.05844710932837593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,32,0.05904533465703329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,51200,1.1078177558051217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,65536,1.392730712890625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,16384,1.0888195037841797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,51200,2.876286188761393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,16384,1.1650435129801433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,65536,3.4729120466444225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,16384,0.29792621400621205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,12288,0.751746654510498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,16384,0.9726684358384875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,51200,3.169113794962565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,12288,0.7539128727383084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,12288,0.26969067255655926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,51200,4.515242682562934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,12288,0.6460364659627279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,10240,0.5820604430304633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,8192,0.4479751057094998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,10240,0.6383857727050781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,65536,4.055334303114149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,8192,0.4937564531962077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,10240,0.5274977684020996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,7168,0.416793770260281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,65536,5.8387849595811625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,7168,0.13983200656043157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,8192,0.5663599967956543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,6144,0.36860089831882054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,7168,0.5672275755140517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,10240,0.19474844137827554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,7168,0.37556176715426975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,6144,0.12052178382873535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,6144,0.3808933364020453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,5120,0.3405173354678684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,5120,0.10626044538285996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,4096,0.23193067974514434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,6144,0.32829509841071236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,5120,0.30077865388658315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,3584,0.19421333736843535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,4096,0.08814222282833523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,4096,0.24316533406575522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,5120,0.2772853374481201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,3072,0.14983999729156494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,3584,0.08820622497134739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,3584,0.20760267310672334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,2560,0.13795910941229925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,3072,0.07024621963500977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,3584,0.19798489411671957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,3072,0.18432444996303984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,2048,0.11331466833750407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,2560,0.061724444230397545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,3072,0.17669333351982963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,2560,0.15905688868628606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,1536,0.08465333117379083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,2560,0.15028533670637342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,2048,0.1270337767071194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,8192,0.16455199983384874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,1024,0.06604711214701335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,2048,0.12384622626834446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,1536,0.10116533438364665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,1536,0.10171733299891154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,1024,0.037124445041020714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,768,0.04751911097102695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,4096,0.32120800018310547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,1024,0.07777866390016344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,512,0.03897155655754937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,1024,0.077437334590488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,768,0.034380445877710976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,768,0.06195111407174004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,256,0.02477688921822442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,512,0.031574222776624895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,512,0.05291733476850721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,768,0.06622044245402019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,128,0.022679999470710754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,2048,0.052538666460249156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,256,0.028452446063359577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,256,0.04765689041879442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,512,0.05343022280269199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,64,0.01920622256067064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,32,0.019320888651741874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,128,0.02777066661251916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,128,0.04488977789878845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,256,0.04289422101444668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,1536,0.04368444283803304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,64,0.04554311103290982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,128,0.032884445455339216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,32,0.046168890264299184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,51200,0.9228230582343208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,65536,1.170080926683214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,16384,1.0091155370076497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,51200,2.4791022406684027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,16384,0.9860507117377387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,16384,0.25629867447747123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,65536,3.2222338782416453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,12288,0.6833759943644205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,51200,2.7073927985297312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,16384,0.8079893324110242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,12288,0.6251840061611599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,12288,0.2248622311486138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,12288,0.5557742118835449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,10240,0.5062382486131456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,51200,4.19603008694119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,10240,0.19382311238182914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,65536,3.456342273288303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,8192,0.41858310169643825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,10240,0.5771546893649632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,8192,0.1567013396157159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,10240,0.457381354437934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,7168,0.3603991137610541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,7168,0.1263964441087511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,8192,0.49268447028266055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,8192,0.3739502165052626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,6144,0.25794490178426105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,65536,5.559606340196397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,7168,0.5253333515591091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,5120,0.24215910169813368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,7168,0.3241742187076145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,5120,0.11805244286855061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,6144,0.28466134601169163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,5120,0.2687137656741672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,3584,0.16073777940538195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,4096,0.07648444175720215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,4096,0.2062293291091919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,5120,0.24645421240064833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,3584,0.06692444615893893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,3072,0.14407644006941053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,4096,0.19444799423217773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,3584,0.17874933613671196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,3072,0.05990311172273424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,2560,0.10819466908772786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,3584,0.17098132769266763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,6144,0.1042062242825826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,3072,0.1645466619067722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,2048,0.10463111268149482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,2560,0.05333866675694784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,3072,0.1507386631435818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,2560,0.13185244136386448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,1536,0.07445600297715929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,4096,0.1849671072430081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,6144,0.31938221719529897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,2048,0.1112755537033081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,1024,0.054971555868784584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,2560,0.1288479963938395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,1536,0.03824355536037021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,1536,0.08646666341357762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,2048,0.11009155379401313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,768,0.039790223042170204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,1024,0.03247377938694424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,1536,0.08819733063379924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,1024,0.06673955255084567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,512,0.03256444467438592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,768,0.030307554536395605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,1024,0.06815733512242635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,256,0.022038223014937505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,512,0.028115553988350764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,512,0.04530310961935255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,768,0.05827999777264065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,128,0.01870577699608273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,256,0.02476177778508928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,512,0.047998223039839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,256,0.040066666073269315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,64,0.01661600006951226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,128,0.03819288810094198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,128,0.023927999867333308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,32,0.017536888519922893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,64,0.03853333327505324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,128,0.028735111157099407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,32,0.03915111223856608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,768,0.05334488881958855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,2048,0.04538044333457947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,256,0.0372977786593967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,65536,0.9664186901516385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,51200,0.7442062166002062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,16384,0.916568014356825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,65536,2.5699689653184676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,51200,2.3359243604871964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,16384,0.20597244633568657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,16384,0.7500969039069282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,12288,0.542468441857232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,16384,0.615415096282959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,51200,3.0154658423529734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,65536,2.765442742241753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,12288,0.16346844037373862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,65536,3.6903368631998696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,10240,0.13468978140089247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,10240,0.39933866924709743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,51200,2.17831908331977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,12288,0.44875113169352215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,8192,0.32504089673360187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,10240,0.43077956305609805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,8192,0.12359289328257243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,7168,0.2869315677218967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,10240,0.3795635435316298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,8192,0.3280071152581109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,7168,0.30815288755628795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,7168,0.10765777693854438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,8192,0.3046240011850993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,7168,0.25932266977098256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,12288,0.5115662150912815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,5120,0.21098132928212485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,6144,0.2565751075744629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,6144,0.08584800031450059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,4096,0.1547902160220676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,5120,0.07404800256093343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,6144,0.2253911230299208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,5120,0.21375021669599745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,3584,0.14314044846428767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,4096,0.06831200255288018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,4096,0.1653964387045966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,5120,0.19106133778889975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,4096,0.1619902186923557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,3584,0.14413421683841282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,2560,0.0953857766257392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,3072,0.049508445792728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,3584,0.14052089055379233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,3072,0.1263128916422526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,2048,0.07482666439480253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,6144,0.27580444018046063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,2560,0.04423288835419548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,3072,0.1248391071955363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,2560,0.10842666361067031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,1536,0.06450755728615655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,2048,0.037885334756639265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,2048,0.08979644378026326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,2560,0.1071831120385064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,1024,0.03973066806793213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,1536,0.03177155719863044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,1536,0.07112533516354032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,2048,0.0902435580889384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,768,0.031197332673602637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,1024,0.025593777497609455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,1024,0.05429511268933614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,1536,0.07418577538596259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,512,0.02384088933467865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,3072,0.1193315585454305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,3584,0.056335111459096275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,768,0.02405066622628106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,768,0.04383022255367703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,256,0.01793955597612593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,1024,0.05836088789833916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,512,0.02160533269246419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,768,0.048467556635538735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,512,0.03570222192340427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,256,0.01957155598534478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,128,0.014018666413095264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,64,0.013100444442696042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,256,0.031165334913465712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,512,0.04040177663167318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,32,0.01384800010257297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,128,0.01976799964904785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,256,0.031832890378104314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,128,0.029842668109469946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,64,0.029509332444932725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,128,0.024679111109839544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,32,0.02987733483314514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,65536,0.9172880384657117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,51200,1.7399813334147136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,65536,2.07206302218967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,51200,0.7623706923590766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,16384,0.5828080177307129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,16384,0.23798934618631998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,16384,0.6225191222296821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,51200,1.9673635694715712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,12288,0.4103591177198622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,16384,0.5557129118177626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,12288,0.15961510605282253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,65536,3.355091518825955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,65536,2.5807776980929904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,10240,0.38045512305365664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,12288,0.55203644434611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,10240,0.12573956118689641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,10240,0.39214666684468585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,8192,0.28776799307929146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,12288,0.40933778550889754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,8192,0.10123555527793036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,7168,0.21957955095503065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,10240,0.34568267398410374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,8192,0.4406604501936171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,7168,0.09077599975797866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,51200,2.676413430107964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,6144,0.18712977568308511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,7168,0.315120882458157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,8192,0.27462132771809894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,6144,0.07992177539401583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,5120,0.17146400610605875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,7168,0.2405137750837538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,6144,0.2229315439860026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,4096,0.13859732945760092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,5120,0.06803644365734525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,5120,0.18992088900672066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,3584,0.11462222205268012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,4096,0.05719821982913547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,6144,0.2273555596669515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,4096,0.14662133322821722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,5120,0.18034933673010933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,3584,0.051495111650890775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,3584,0.12846222188737658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,4096,0.1459724505742391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,3072,0.04662400152948168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,3072,0.11239733960893418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,3584,0.1308142211702135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,2048,0.06752622127532959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,2560,0.03997688823276096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,2560,0.09674666987525092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,3072,0.1158391104804145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,1536,0.055008888244628906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,2048,0.03380266825358073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,2560,0.09887733062108357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,2048,0.08021333482530382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,1024,0.0387715564833747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,1536,0.06417777803209093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,1536,0.02866311205757989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,768,0.02888266576661004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,2048,0.08364533053504096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,1024,0.024320888850424025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,1536,0.06780622402826945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,1024,0.04829688866933187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,512,0.020756444997257657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,768,0.022090666823916968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,1024,0.05300888750288221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,768,0.03968800107638041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,256,0.015399111641777886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,512,0.03170399864514669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,512,0.019495111372735765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,768,0.04504266712400648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,128,0.0144213338692983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,256,0.02757688860098521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,512,0.03784444597032335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,64,0.012404444317022959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,128,0.017940445078743827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,128,0.026605332891146343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,32,0.012425777812798819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,128,0.02233244478702545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,64,0.026900443765852187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,32,0.02664088871743944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,3072,0.10381599929597642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,2560,0.09064533313115437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,256,0.018039110634062026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,256,0.029312888781229656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,65536,0.7636382314893934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,51200,0.5883199903700087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,51200,1.5178764131334093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,65536,2.2925056881374783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,16384,0.6496008767022027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,16384,0.5397200054592556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,65536,2.205615997314453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,51200,2.213782204522027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,16384,0.18955111503601074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,16384,0.4973306655883789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,51200,1.7004746331108942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,12288,0.33728355831570095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,12288,0.12645688321855333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,10240,0.29527023103502065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,12288,0.44360799259609646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,10240,0.10793511072794597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,8192,0.24220445421006945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,10240,0.40741244951883954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,12288,0.35862310727437335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,8192,0.1017635530895657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,7168,0.2190586725870768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,10240,0.2950879997677273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,8192,0.25100355678134495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,65536,2.9182559119330516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,6144,0.18559555212656656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,7168,0.2266613377465142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,8192,0.23863199022081164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,6144,0.0693004462454054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,6144,0.1894435617658827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,7168,0.21697244379255506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,5120,0.0596275528271993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,4096,0.12008978260887994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,5120,0.16015377309587267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,3584,0.09860266579522027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,4096,0.04972444309128655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,5120,0.15518400404188368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,4096,0.13050044907463923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,3584,0.04451910985840691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,3072,0.08415200312932332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,4096,0.1295022169748942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,3584,0.11366932921939427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,2560,0.07071555323070951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,3072,0.03973689013057285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,3584,0.11463911003536648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,7168,0.07883199718263414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,2048,0.06392355759938557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,2560,0.034817778401904635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,5120,0.13563911120096842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,3072,0.10116266542010838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,1536,0.04273600048489041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,2048,0.030049777693218652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,6144,0.18376889493730333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,2560,0.08842221895853679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,1024,0.030935999419954088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,1536,0.05580355723698934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,1536,0.025109334124459162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,2048,0.07393422391679552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,768,0.023090667194790308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,1024,0.042908443344963916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,1536,0.06040800280041165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,768,0.03485511077774896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,768,0.01943733294804891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,1024,0.04721600148412916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,512,0.017086222767829895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,3072,0.09882133536868626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,256,0.013361777696344586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,512,0.017412443955739338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,768,0.04082666503058539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,512,0.028189334604475234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,2560,0.08508444494671291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,128,0.010642666783597736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,256,0.016187555260128446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,512,0.033583998680114746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,64,0.009581333233250512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,2048,0.07103911373350355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,128,0.016192000773217943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,32,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,256,0.026280888252788123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,128,0.020053333706325956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,64,0.0233608881632487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,32,0.023375110493765935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,1024,0.021537777450349595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,65536,0.6608631345960829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,256,0.024412444896168176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,128,0.023344000180562336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,65536,1.5995261934068468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,51200,1.3002782397800021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,16384,0.4021004570855035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,51200,0.5097706582811143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,16384,0.14943733480241564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,12288,0.33831556638081867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,16384,0.4745573467678494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,51200,1.9093360900878906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,16384,0.4616186883714464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,12288,0.11151022381252712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,65536,2.6465706295437283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,12288,0.34219378895229763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,10240,0.2735564443800184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,65536,1.9252630869547527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,8192,0.20585866769154867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,10240,0.10695111089282566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,10240,0.2769182258182102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,12288,0.37093954616122776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,8192,0.07867733637491862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,51200,1.498313797844781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,10240,0.2674204508463542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,8192,0.22628000047471789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,6144,0.1568382183710734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,7168,0.07887911134295993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,7168,0.19559200604756674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,6144,0.06473510795169406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,5120,0.12063111199273004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,7168,0.19671911663479277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,6144,0.16280889511108398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,5120,0.0533591111501058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,4096,0.08946666452619766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,6144,0.16689688629574248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,5120,0.1402088933520847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,4096,0.043839110268486865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,5120,0.14220710595448813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,4096,0.11056177483664618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,8192,0.23099644978841147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,7168,0.1821155548095703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,3072,0.08131200075149536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,3584,0.03956088754865859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,3584,0.10003555483288235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,2560,0.05613955524232653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,3072,0.03471022182040744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,3584,0.10456177923414443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,3072,0.08584888776143391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,2048,0.05479377839300367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,2560,0.03091555502679613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,2560,0.07366400294833712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,1536,0.037216001086764865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,2048,0.026123555170165166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,2560,0.08124088578754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,2048,0.06122844749026828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,1024,0.02564977771706051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,1536,0.048822220828798085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,3584,0.08639644251929389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,2048,0.06765421893861559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,768,0.020429333051045735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,1024,0.018686221705542672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,1536,0.05498844385147095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,4096,0.11749777528974746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,512,0.015992000699043274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,768,0.017571555243598092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,768,0.030563556485705908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,512,0.01615111033121745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,768,0.036904888020621404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,512,0.024496000674035814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,256,0.01262399968173769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,3072,0.09224800268809001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,128,0.009874666730562845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,256,0.015062222878138224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,512,0.030784000953038532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,256,0.021312000022994146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,64,0.010648888846238455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,128,0.014818666709793938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,256,0.023261333505312603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,128,0.020459555917316012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,32,0.011503111157152386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,128,0.018252443936136033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,64,0.0203457772731781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,1024,0.037288000186284385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,32,0.020273778173658583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,1024,0.04370489054256015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,65536,0.5436551305982802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,51200,0.9632017347547742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,1536,0.022359111242824133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,65536,1.4095635943942602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,16384,0.3261920081244575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,51200,0.4609164661831326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,16384,0.1255697806676229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,16384,0.3657795588175456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,51200,1.4179475572374132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,12288,0.28296176592508954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,65536,1.9354542626274958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,51200,1.2076311111450195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,12288,0.09544533491134644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,16384,0.3698604371812608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,10240,0.24675199720594618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,65536,1.6128693686591253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,12288,0.27488801214430064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,10240,0.08192711406283908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,8192,0.19546577665540907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,10240,0.23417867554558647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,12288,0.26913245519002277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,7168,0.14241688781314427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,8192,0.06496355268690321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,8192,0.24581866794162324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,10240,0.23411732249789766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,6144,0.13665333059098986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,7168,0.05821600225236681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,8192,0.19056533442603218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,7168,0.1784364382425944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,5120,0.09913866387473212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,6144,0.05169066786766052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,6144,0.13620711697472468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,4096,0.08538311057620579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,7168,0.16339022583431667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,5120,0.047988444566726685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,5120,0.1147280004289415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,3584,0.07707289192411634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,4096,0.036438223388459944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,6144,0.14562133948008218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,4096,0.09333244297239517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,5120,0.12160178025563557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,3584,0.03229511115286086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,3584,0.08180266618728638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,4096,0.09985333018832737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,2560,0.05739733245637682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,3072,0.02908710969818963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,3072,0.07125600179036458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,3584,0.08883821964263916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,2048,0.04592622319857279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,2560,0.026241777671708003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,2560,0.0616133345497979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,3072,0.07897244559393989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,1536,0.035798221826553345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,2048,0.022612444228596155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,2048,0.05130844314893087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,2560,0.06906400124231975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,1024,0.02534577747186025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,1536,0.01902844508488973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,1536,0.040658666027916804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,2048,0.05886399745941162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,768,0.01756355497572157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,1024,0.01635199950801002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,1024,0.031069331698947485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,1536,0.04810666706826952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,512,0.014801777071423002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,3072,0.06341155370076497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,768,0.015461333923869662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,256,0.011686222420798408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,1024,0.03725333346260919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,512,0.014136888914638095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,768,0.03147466646300422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,128,0.007699555820888943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,256,0.01330044451687071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,256,0.01791288952032725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,64,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,128,0.013268444273206921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,32,0.0069848886794514125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,128,0.017516444126764934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,128,0.017329777280489605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,64,0.017417778571446735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,32,0.01723022262255351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,768,0.025976000560654536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,512,0.021140444609853957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,65536,0.4642008675469293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,65536,1.0099670622083876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,512,0.02574577762020959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,256,0.021313778228229944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,16384,0.28151022063361275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,51200,0.7471289104885526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,51200,0.3560009002685547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,16384,0.11150844891866048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,12288,0.20677688386705187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,16384,0.3216133382585314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,65536,1.6074355443318684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,65536,1.4069102605183919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,10240,0.17718311150868735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,12288,0.08500177992714776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,16384,0.3282151222229004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,51200,1.2784710990058052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,10240,0.07270577881071302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,8192,0.14626044697231716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,12288,0.24933332867092559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,51200,1.0987573199801974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,10240,0.24260089132520887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,8192,0.06096000141567654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,7168,0.12480710612403022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,10240,0.2284151183234321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,8192,0.1787733369403415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,7168,0.052730666266547314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,6144,0.09522489044401382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,5120,0.08424711227416992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,7168,0.17009511258867052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,6144,0.046282665597067944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,6144,0.12488267156812881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,7168,0.15065866046481663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,5120,0.038894222842322454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,4096,0.06857866711086698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,5120,0.108897778722975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,6144,0.13161688380771214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,3584,0.06460177898406982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,4096,0.03221688999070062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,4096,0.08356089062160915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,5120,0.10980888870027329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,12288,0.29669777552286786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,3072,0.05340977840953403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,3584,0.029270221789677937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,3584,0.07441066371070014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,4096,0.09143111440870498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,3072,0.02643111182583703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,3072,0.06453422043058607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,3584,0.08153333266576131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,2048,0.03695111142264472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,2560,0.055267555846108325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,2560,0.023872888750500146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,3072,0.07284533315234713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,1536,0.02683199942111969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,2048,0.021056888831986323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,2560,0.06326399909125434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,2048,0.04652000135845608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,1024,0.01867022282547421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,1536,0.018519111805491976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,1536,0.037295109695858426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,1024,0.02775466607676612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,1024,0.015557333827018738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,768,0.015239111251301236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,1536,0.044060445494121976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,1024,0.03360266817940606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,512,0.011385777758227454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,2560,0.04318044583002726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,768,0.023349333140585158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,768,0.02885422110557556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,512,0.013458666702111563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,256,0.008664888640244802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,512,0.01885777711868286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,8192,0.16453688674502903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,256,0.012599111431174807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,512,0.023927999867333308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,128,0.007328888608349695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,256,0.016297777493794758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,64,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,128,0.012424888710180918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,32,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,2048,0.05291555656327141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,128,0.01585244470172458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,64,0.01552088889810774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,128,0.01588000026014116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,32,0.015915556086434256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,768,0.01481244464715322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,65536,0.42022045453389484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,65536,1.0152568817138672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,51200,0.8327386644151477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,256,0.02127644419670105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,51200,0.3369013203514947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,16384,0.28914043638441295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,65536,1.3611919615003798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,16384,0.10785600211885239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,51200,1.0112071567111545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,12288,0.2113493283589681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,16384,0.3741724491119385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,65536,1.219552887810601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,10240,0.17676799827151826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,12288,0.08244799905353122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,16384,0.28554222318861217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,12288,0.22838134235805937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,8192,0.14247643947601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,10240,0.07383733325534396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,10240,0.18245689074198404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,7168,0.11190756162007649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,12288,0.23553511831495497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,10240,0.19741511344909668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,7168,0.04637688729498121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,6144,0.10045688682132298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,8192,0.1495306624306573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,7168,0.12752089235517713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,5120,0.08375377787484063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,6144,0.040692445304658674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,7168,0.13185155391693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,6144,0.10853066709306504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,4096,0.06573510832256742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,5120,0.03418488966094123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,6144,0.11537244584825303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,5120,0.0930773350927565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,51200,0.9286631478203667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,3584,0.05382844474580553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,4096,0.029095109966066148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,5120,0.09870044390360515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,4096,0.0738933351304796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,3584,0.025994666748576697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,3584,0.06528088781568739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,8192,0.05503911111089918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,4096,0.08158755302429199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,2560,0.040399110979504056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,8192,0.1431822246975369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,3584,0.07179733117421468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,3072,0.023559111687872145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,2048,0.03267466690805223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,2560,0.021677333447668288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,2560,0.049512889650132924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,3072,0.06530933247672187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,1536,0.024919999970330134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,2048,0.01905777719285753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,2048,0.041839109526740186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,2560,0.056185775332980685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,1024,0.016729777057965595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,1536,0.016755556066830952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,2048,0.047572443882624306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,1536,0.03335377905103896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,768,0.013904000322024027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,1024,0.014373333917723762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,1024,0.024949333733982507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,512,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,1536,0.038880000511805214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,1024,0.03012088934580485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,3072,0.047116445170508496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,768,0.021007110675175984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,256,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,512,0.012548444171746572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,3072,0.057067553202311196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,512,0.021661332911915247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,128,0.006575999988449945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,256,0.01219288922018475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,256,0.018592889110247295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,128,0.014398222168286642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,64,0.006135111053784688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,128,0.011706666813956367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,32,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,128,0.014925332532988654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,64,0.013982221484184265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,768,0.013478221992651621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,32,0.01442577772670322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,512,0.017222222354676988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,256,0.014818666709793938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,65536,0.36700977219475645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,65536,0.9235191345214844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,51200,0.5717413160536025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,16384,0.21277777353922525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,65536,1.1198648876614041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,51200,0.3394808769226074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,51200,0.850136015150282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,768,0.02588355541229248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,16384,0.10977777507570054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,12288,0.16388621595170763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,65536,1.1382630666097004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,16384,0.3088737858666314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,10240,0.11948800086975098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,12288,0.08191644483142428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,12288,0.19023466110229492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,51200,0.8974747127956815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,8192,0.11027466588550144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,10240,0.061215996742248535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,10240,0.16140088770124647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,7168,0.09448444181018406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,12288,0.2116560008790758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,8192,0.049739556180106274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,10240,0.18417066997951934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,8192,0.12573956118689641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,7168,0.04289066791534424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,7168,0.11255644427405463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,8192,0.149834672609965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,5120,0.06140266524420845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,6144,0.03707111212942336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,7168,0.13047199779086643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,4096,0.05225333240297106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,5120,0.032023111979166664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,5120,0.08123199807273017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,6144,0.11246221595340306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,3584,0.04488089018397861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,5120,0.09543733464346991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,4096,0.027113777067926195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,16384,0.2870417700873481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,3072,0.0383857786655426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,3584,0.024704888463020325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,4096,0.07854933208889432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,3584,0.05771911144256592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,6144,0.077638222111596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,2560,0.030064887470669214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,3584,0.07049600283304851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,3072,0.0510791109667884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,2048,0.028149333265092637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,2560,0.02058844433890449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,3072,0.06291733185450236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,6144,0.09572800000508626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,2560,0.04412800073623657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,1536,0.02055022286044227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,2048,0.036869333850012884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,2048,0.017680888374646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,2560,0.055120888683531016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,1024,0.015035554766654968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,1536,0.01553244392077128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,1536,0.02955289019478692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,768,0.011209777659840055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,2048,0.04504799842834473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,4096,0.06525778108172946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,1024,0.02232711182700263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,1024,0.013463999662134381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,1536,0.03825777769088745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,512,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,768,0.013179555535316467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,1024,0.02945599953333537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,768,0.018723555737071566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,512,0.012109333442317115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,512,0.015454222758611044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,128,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,256,0.011384000380833944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,512,0.020768000019921195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,128,0.01143022212717268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,256,0.017310221989949543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,128,0.01277511070171992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,32,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,128,0.014325333966149224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,64,0.012517333030700684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,32,0.01272266689274046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,256,0.006903999795516332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,65536,0.6078222062852647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,768,0.025019556283950806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,256,0.013343111508422427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,65536,0.3141075505150689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,64,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,65536,0.8753537601894803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,51200,0.45174487431844074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,16384,0.17474577162000868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,51200,0.24662934409247506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,65536,0.9936569001939562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,16384,0.21033244662814668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,51200,0.6769955423143176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,16384,0.08063022295633952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,12288,0.13287022378709581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,10240,0.1200631062189738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,12288,0.06258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,3072,0.02218755582968394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,16384,0.2514817714691162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,12288,0.15679466724395752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,8192,0.09123377667533027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,51200,0.7752328978644477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,10240,0.05333244469430712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,10240,0.14080711205800375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,12288,0.19094844659169516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,8192,0.04797155658404032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,8192,0.10611555311414932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,10240,0.1575511164135403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,6144,0.06840533018112183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,7168,0.03943022092183431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,7168,0.0975831084781223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,8192,0.1284159951739841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,5120,0.057074666023254395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,6144,0.0332622230052948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,7168,0.11343377166324192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,6144,0.08204711145824857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,4096,0.04341244366433886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,5120,0.06866844495137532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,3584,0.04075377848413255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,6144,0.09813777605692546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,4096,0.05471200082037184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,5120,0.08409244484371609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,3072,0.03359466791152954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,4096,0.07020088699128893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,3584,0.021944888763957556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,3584,0.048736890157063804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,2560,0.029517332712809246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,7168,0.08027555545171101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,3072,0.019928889142142404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,3584,0.06271911329693265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,3072,0.043271111117468945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,2560,0.018208000395033095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,2560,0.037792887952592634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,3072,0.05633422401216295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,1536,0.01924622224436866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,2048,0.031607998741997614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,2048,0.01626755628320906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,2560,0.0479928884241316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,1024,0.013869333598348828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,1536,0.014828445182906257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,1536,0.026135999295446608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,2048,0.04121422105365329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,5120,0.02836444311671787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,768,0.01164533363448249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,4096,0.024129778146743774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,1024,0.019555555449591745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,1024,0.012898666991127862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,512,0.008637333081828224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,1536,0.03401155604256524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,768,0.012048888537618848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,768,0.016530666086408827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,256,0.006221333311663733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,512,0.0110871113008923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,1024,0.026829333768950567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,512,0.013743999931547375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,768,0.022963555322753057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,128,0.006081777728266186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,256,0.010692444112565784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,256,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,2048,0.02345955537425147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,64,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,512,0.018903111418088276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,128,0.01073333372672399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,128,0.011424000064531961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,256,0.015757333901193406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,64,0.011060444845093621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,128,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,32,0.011083555718262991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,65536,0.2620026535458035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,51200,0.3599084483252631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,16384,0.1325733396742079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,65536,0.686157332526313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,51200,0.20786400636037192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,51200,0.5368328624301487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,16384,0.07527199718687269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,12288,0.10004622406429714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,16384,0.18158222569359672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,32,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,51200,0.6255448659261068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,10240,0.07897244559393989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,12288,0.058119111590915255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,16384,0.20931022697024873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,12288,0.13268621762593588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,8192,0.06643555561701457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,10240,0.04729511009322273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,10240,0.11004444625642563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,12288,0.1602097749710083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,65536,0.5145413080851237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,7168,0.05834311246871948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,8192,0.035086221165127225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,10240,0.13507289356655544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,8192,0.0883057779735989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,6144,0.0474275549252828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,7168,0.03150577677620782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,7168,0.07771288686328463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,5120,0.039020445611741804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,8192,0.11167111661699082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,6144,0.027628444963031348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,6144,0.06763466861512926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,7168,0.09762489133410984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,4096,0.029086222251256306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,65536,0.8066275384691027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,6144,0.08449421988593207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,5120,0.056537780496809215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,3584,0.02684266699684991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,4096,0.04641333222389221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,4096,0.020392888122134738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,5120,0.07310133510165744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,3072,0.022351110974947613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,3584,0.04192089041074117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,3584,0.018932445181740653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,4096,0.06144622299406263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,2560,0.0196435550848643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,3072,0.03707733419206407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,3072,0.01718222267097897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,3584,0.05467199948098925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,2048,0.016120000018013846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,2560,0.031940443648232356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,2560,0.01607111096382141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,3072,0.049287110567092896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,1536,0.0127697777416971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,2048,0.014460444450378418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,2048,0.02711733347839779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,2560,0.041802667909198336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,1024,0.009658666948477427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,1536,0.022099556194411382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,1536,0.013075555364290873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,2048,0.03496799866358439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,768,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,5120,0.02407377792729272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,1024,0.011744000017642975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,1536,0.02915999955601162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,1024,0.01684711045689053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,512,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,256,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,768,0.010960000256697336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,768,0.01443288889196184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,1024,0.023752000596788194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,512,0.010330666270520952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,128,0.004250666747490565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,512,0.01184266640080346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,256,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,256,0.010466666685210334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,64,0.004053333153327306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,32,0.0041964443193541635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,512,0.017808000246683758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,128,0.0099973330895106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,128,0.010103999740547603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,256,0.014811555544535318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,128,0.012064889073371887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,64,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,32,0.009656888743241629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,65536,0.32860000928243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,65536,0.2035964462492201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,51200,0.2567395634121365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,16384,0.09205866522259182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,65536,0.5380737516615126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,51200,0.16344443957010904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,12288,0.07018578052520752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,16384,0.05557866891225179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,768,0.02026844521363576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,65536,0.6543404261271158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,10240,0.05927111042870415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,12288,0.044009778234693736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,12288,0.1334542168511285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,10240,0.0879395537906223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,8192,0.046662221352259316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,10240,0.038650665018293597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,10240,0.11292889383104111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,8192,0.02806666824552748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,8192,0.07075733608669706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,7168,0.040299554665883384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,16384,0.13829599486456975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,8192,0.09336177508036296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,7168,0.02565866708755493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,6144,0.03380977776315477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,51200,0.5075688891940647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,7168,0.06234221988254123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,5120,0.028352889749738906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,16384,0.17264178064134386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,6144,0.0221395558781094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,7168,0.08157955275641547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,6144,0.05453777644369337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,4096,0.023818666736284893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,5120,0.019547555181715224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,5120,0.0459191103776296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,6144,0.07139999998940362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,3584,0.021344888541433547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,4096,0.016972444123691983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,4096,0.03849066628350152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,3584,0.034653332498338484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,4096,0.05072089036305746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,3584,0.015981333123313058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,51200,0.4179840087890625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,3584,0.04598844382497999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,2560,0.016473778420024447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,3072,0.014841778410805596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,3072,0.030584888325797185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,12288,0.10708800289365981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,2048,0.013778666655222574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,3072,0.04090311129887899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,2560,0.02648711038960351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,2048,0.01240711079703437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,2560,0.03571288784344991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,2048,0.022208889325459797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,1536,0.01108088923825158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,5120,0.061216890811920166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,2048,0.030099554194344416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,1024,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,1536,0.01851377718978458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,3072,0.01867911054028405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,1536,0.024872000018755596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,1024,0.010355555348926121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,1024,0.014156444205178155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,768,0.01000088867213991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,1024,0.02032533288002014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,512,0.005882666756709416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,768,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,256,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,512,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,2560,0.013697778185208639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,768,0.01793066660563151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,512,0.01033688916100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,128,0.004200889004601372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,256,0.008968888885445064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,256,0.00869599978129069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,512,0.015454222758611044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,128,0.008973333570692275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,32,0.0039057777159743835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,256,0.013442666994200813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,1536,0.011345778074529437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,128,0.010673777924643623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,64,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,32,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,768,0.007352888584136963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,65536,0.19816266165839302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,51200,0.20805511209699842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,65536,0.4528515603807237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,16384,0.06819911135567559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,51200,0.36039021280076766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,64,0.003908444609906939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,65536,0.6479893260531956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,128,0.008053333395057255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,16384,0.11801777945624457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,12288,0.051670223474502563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,16384,0.05470755696296692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,51200,0.5059546894497341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,12288,0.04334577918052673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,16384,0.17131377591027153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,65536,0.25783467292785645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,8192,0.03327466713057624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,10240,0.03775022096104092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,10240,0.0750088890393575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,12288,0.13212622536553278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,7168,0.028103109863069322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,8192,0.027830223242441814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,8192,0.06076888905631172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,10240,0.11205511622958714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,6144,0.023988443944189284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,7168,0.02495555579662323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,7168,0.053673777315351695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,8192,0.09177066882451375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,5120,0.01907911068863339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,6144,0.022044445077578228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,6144,0.04671555426385668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,7168,0.08026222387949626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,4096,0.015928000211715698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,5120,0.01963911122745938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,5120,0.04015466570854187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,6144,0.07043644454744127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,3584,0.014512888259357877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,12288,0.08933866686291164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,4096,0.017204445269372728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,4096,0.03282399972279867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,5120,0.06041244665781657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,3072,0.012643555800120035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,3584,0.015810666812790763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,3584,0.029765334394243028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,2560,0.011000889043013254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,4096,0.05039644572469923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,3072,0.026130666335423786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,3072,0.014817777607176038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,3584,0.045127110348807446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,51200,0.15825777583652073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,2560,0.02276266614596049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,3072,0.03978933228386773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,2048,0.01241422196229299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,2560,0.03464533223046197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,2048,0.019280888968043856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,1536,0.008014221986134848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,1536,0.011387555963463254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,10240,0.04251111216015286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,2048,0.02939644455909729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,1024,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,1536,0.016192888220151264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,768,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,1024,0.010355555348926121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,1536,0.02437866727511088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,1024,0.012960889273219638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,512,0.004401777767472797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,768,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,1024,0.01997777819633484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,512,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,512,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,2048,0.00960177762640847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,256,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,2560,0.013471110827393003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,128,0.0031662223239739737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,256,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,512,0.015111999379263984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,256,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,64,0.002875555513633622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,128,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,128,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,32,0.0031715554909573663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,256,0.013194666968451606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,64,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,128,0.01036533299419615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,65536,0.20372445053524443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,32,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,51200,0.1639288928773668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,65536,0.15244978004031712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,768,0.010837333069907295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,65536,0.40185777346293133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,16384,0.0522017776966095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,51200,0.12250844637552898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,51200,0.31446043650309247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,12288,0.038626667526033186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,16384,0.04294666647911072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,16384,0.10555111037360297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,10240,0.032398223876953125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,65536,0.5050684611002604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,12288,0.03439022103945414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,16384,0.13596799638536242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,51200,0.4031546645694309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,8192,0.024273778001467388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,10240,0.029951999584833782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,12288,0.10609689023759629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,10240,0.06744888756010267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,7168,0.02083288960986667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,8192,0.02125155594613817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,10240,0.08973511060078938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,6144,0.018391110830836825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,7168,0.018976888722843595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,7168,0.04844889044761658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,8192,0.07295111152860853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,5120,0.015863110621770222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,6144,0.017050666941536795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,7168,0.06416355239020453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,6144,0.04206933246718513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,768,0.017895999881956313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,4096,0.013424889081054263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,5120,0.015184000134468079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,5120,0.035994665490256414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,3584,0.01163999984661738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,4096,0.013398222625255585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,4096,0.029655112160576716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,5120,0.048551999860339694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,3072,0.010685332947307162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,4096,0.04007200068897671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,3584,0.026498667067951624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,3072,0.012051555845472546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,3072,0.023894222246276006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,3584,0.03564888901180691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,8192,0.054641776614718966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,3072,0.031862222486072116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,2560,0.011362666885058085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,2048,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,2560,0.02085777786042955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,2048,0.010743111371994019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,2048,0.017835555805100333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,6144,0.056571556462181934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,2048,0.024365334047211543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,1536,0.014658666319317289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,12288,0.08028888702392578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,1536,0.00997511131895913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,1024,0.005208000126812193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,3584,0.01258933295806249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,768,0.004678222040335338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,1024,0.009312888814343346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,1536,0.020615999897321064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,1024,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,2560,0.009492444495360056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,768,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,512,0.003896000070704354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,1024,0.01720888912677765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,768,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,256,0.0035048888789282907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,512,0.00830844458606508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,768,0.014997333288192749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,512,0.008669333325492011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,128,0.003539555602603488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,256,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,256,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,1536,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,64,0.0032213332338465583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,128,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,256,0.011719110939237805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,32,0.0033404444240861465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,65536,0.11946045027838813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,64,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,128,0.0090515555606948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,32,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,51200,0.09388799799813165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,65536,0.15166933006710476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,65536,0.32730044258965385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,51200,0.25697867075602215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,51200,0.1212115552690294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,2560,0.02829688787460327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,512,0.01313777764638265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,65536,0.4873555501302083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,12288,0.027289777994155884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,128,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,51200,0.3818044397566054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,16384,0.08597333563698663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,10240,0.021174222230911255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,12288,0.06476710902320014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,12288,0.03392088744375441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,16384,0.13231289386749268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,8192,0.016394666499561734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,10240,0.029285334878497656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,10240,0.054660442802641124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,12288,0.10233155886332194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,7168,0.014700444208251106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,8192,0.02040177749262916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,8192,0.04426755507787069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,10240,0.08673599693510269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,6144,0.013004444539546967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,7168,0.018572444717089336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,7168,0.039593779378467135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,8192,0.07080800003475614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,5120,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,16384,0.036772443188561335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,6144,0.016748444901572335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,7168,0.06109866831037733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,6144,0.03443200058407254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,4096,0.009384888741705153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,5120,0.015136000182893542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,5120,0.029716445340050593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,6144,0.05420177843835619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,3584,0.008640000389681922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,4096,0.013284444808959961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,3072,0.007865777446164025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,4096,0.024383111132515803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,5120,0.04695555567741394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,3584,0.012578667038016848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,2560,0.006994666324721442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,3584,0.022063111265500385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,4096,0.03833955526351929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,3072,0.012115555504957834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,2048,0.005971555494599872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,3584,0.03455377618471781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,2560,0.011371555427710215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,3072,0.03051999873585171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,1536,0.005192000004980299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,2048,0.010760000182522668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,2048,0.014888889259762235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,2560,0.027469333675172593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,1024,0.0041991112132867174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,1536,0.010049777726332346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,2048,0.023629332582155865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,1536,0.01242844429281023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,768,0.0038604442444112566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,1024,0.00924533357222875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,1024,0.009995555712117089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,1536,0.019909333851602342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,512,0.0031617778456873367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,768,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,768,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,1024,0.016501333978441026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,256,0.0028355556229750314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,512,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,512,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,768,0.014816888504558139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,128,0.0025457777082920074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,3072,0.01962933275434706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,256,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,256,0.00630044440428416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,2560,0.017283555534150865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,512,0.012886222038004132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,64,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,32,0.002208000048995018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,128,0.00795999997191959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,128,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,256,0.011114666859308878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,64,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,65536,0.09051733546786839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,32,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,128,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,51200,0.07243911425272624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,65536,0.1514684491687351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,65536,0.28440978791978627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,16384,0.042381332980261914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,51200,0.12087910705142552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,16384,0.02773688899146186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,12288,0.020589333441522386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,51200,0.22285066710578072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,16384,0.04321600000063578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,16384,0.0748702221446567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,65536,0.48045778274536133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,12288,0.056813332769605845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,12288,0.034627556800842285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,51200,0.37772178649902344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,16384,0.13128444883558485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,12288,0.10075644652048747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,10240,0.03035822179582384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,10240,0.04799733228153653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,7168,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,8192,0.039084444443384804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,10240,0.08578133583068848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,7168,0.03471466567781236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,8192,0.06976977984110515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,6144,0.00980088859796524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,7168,0.06156000163820055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,6144,0.03050400151146783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,6144,0.01756355497572157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,10240,0.015441776977645027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,6144,0.05394133263164096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,8192,0.01275022245115704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,4096,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,5120,0.015740444262822468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,5120,0.026050666968027752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,3584,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,4096,0.013784888717863293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,8192,0.021596444977654353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,5120,0.04651288853751289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,4096,0.021681777305073206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,3072,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,3584,0.012902222573757172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,3584,0.019792889555295307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,4096,0.03746933407253689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,2560,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,7168,0.01925155520439148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,3072,0.012372444073359171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,3072,0.017803554733594257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,2560,0.011471110913488599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,3072,0.030087110069062974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,2560,0.015674667225943673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,1536,0.0041351111398802865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,2048,0.01366488883892695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,2048,0.011065777805116443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,2560,0.027114666170544092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,1024,0.003504000190231535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,2048,0.022992889086405437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,1536,0.010062221851613786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,1536,0.011485333244005838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,768,0.0031724443866146933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,1024,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,1024,0.009369778136412302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,1536,0.019640889432695176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,512,0.0028400001012616684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,768,0.009016888837019602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,768,0.008279999925030602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,1024,0.016194666425387066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,256,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,512,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,512,0.00702044450574451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,768,0.014394667413499622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,128,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,2048,0.004592888885074192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,256,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,256,0.005782222168313132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,64,0.0021342221233579847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,128,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,128,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,5120,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,32,0.0021377777059872947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,256,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,51200,0.06734489070044623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,32,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,16384,0.027822222974565294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,65536,0.2834017806582981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,12288,0.016700444949997794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,51200,0.22239910231696236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,10240,0.012311110893885294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,16384,0.07452799876530965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,8192,0.012210666305489011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,12288,0.05685866541332669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,7168,0.01090311093462838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,10240,0.04776888754632738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,512,0.012127111355463663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,8192,0.038823998636669584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,7168,0.03459644317626953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,5120,0.008095111283991072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,64,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,4096,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,5120,0.025776000486479864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,3584,0.006317333214812809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,4096,0.02163644466135237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,3072,0.005879111174080107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,3584,0.01964000033007728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,2560,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,3072,0.017823111679818895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,2048,0.004567999806669024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,3584,0.03400889039039612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,1536,0.0038888889054457345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,6144,0.009339555270142024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,1024,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,2560,0.015538665983412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,2048,0.01346933344999949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,768,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,512,0.002570666579736604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,1536,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,6144,0.030179556873109605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,65536,0.08678311109542847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,1024,0.009309333231714036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,128,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,64,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,768,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,32,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,512,0.007010666860474481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,256,0.005759999983840519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,128,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,65536,0.08197244670655993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,64,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,51200,0.06576266553666857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,16384,0.02382844520939721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,32,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,12288,0.014956444501876831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,10240,0.01130488928821352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,65536,0.2842133310106066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,16384,0.07443022065692477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,51200,0.22235555118984648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,12288,0.05663377708858914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,8192,0.009870222045315636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,7168,0.009008888569143083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,6144,0.008452444440788692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,10240,0.04789688852098253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,5120,0.010663111176755695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,4096,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,8192,0.03849688834614224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,7168,0.03420089019669427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,3584,0.008055110772450766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,6144,0.030192888445324365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,3072,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,256,0.0024684444069862366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,5120,0.02574399941497379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,2560,0.00646311127477222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,2048,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,4096,0.02163999941613939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,1536,0.0038737778862317405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,1024,0.0032044444233179092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,3584,0.019663110375404358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,768,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,3072,0.01767200065983666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,2560,0.01595288846227858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,2048,0.013635555903116861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,512,0.00253866674999396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,256,0.002191999927163124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,1536,0.011696000066068439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,128,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,64,0.0019466665883858998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,1024,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,32,0.0018657777044508194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,768,0.00815555536084705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,512,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,256,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,128,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,64,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,32,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,16384,1.5549244350857203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,12288,3.3277049594455295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,16384,4.677288055419922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,12288,1.1625324885050456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,16384,5.0229644775390625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,12288,3.812572479248047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,10240,0.9537120395236545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,51200,5.971368153889974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,10240,2.8200026618109812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,8192,0.8537795808580187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,8192,2.252074771457248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,10240,3.176740434434679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,8192,2.33751466539171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,7168,0.6561502350701226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,7168,2.0373378329806857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,6144,0.5903146531846788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,6144,1.8879386054144964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,7168,2.076889673868815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,6144,1.6668693754408095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,5120,0.5002355575561523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,5120,1.724180433485243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,4096,1.1264817979600694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,4096,0.4148000081380208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,5120,1.4934168921576605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,4096,1.1334809197319877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,3584,1.0378843943277996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,3584,0.36100265714857316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,3584,1.0238311555650499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,3072,0.3143893347846137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,3072,0.8828595479329427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,51200,13.869682312011719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,3072,0.8666248851352267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,2560,0.29849423302544487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,2560,0.772508462270101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,2560,0.7387208938598633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,2048,0.2367253303527832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,1536,0.4778435495164659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,2048,0.5865786870320638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,2048,0.6150257852342393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,1536,0.1909093326992459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,1024,0.31481867366366917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,768,0.25390399826897514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,1024,0.1520248916414049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,1536,0.47336000866360134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,768,0.13210399945576987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,512,0.19234044022030303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,768,0.27582044071621364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,1024,0.3895146581861708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,256,0.11792443858252631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,512,0.11913955211639404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,128,0.10138399733437432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,512,0.23703021473354766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,256,0.10359377993477716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,64,0.0856035550435384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,128,0.10181688600116306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,32,0.08668355809317695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,256,0.24952265951368544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,128,0.20190933015611437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,64,0.20429333051045737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,32,0.21013243993123373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,128,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,16384,3.537808100382487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,51200,4.6019672817654085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,51200,22.30196804470486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,65536,6.162935045030381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,16384,1.199262195163303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,16384,3.97255367702908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,12288,0.8424871232774523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,12288,2.6478924221462674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,10240,0.7204693158467611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,12288,2.9660542805989585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,51200,10.971107482910156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,10240,2.577031029595269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,10240,2.4637235005696616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,8192,0.580551094479031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,8192,2.0713378058539496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,8192,1.8625626034206813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,7168,0.5129635598924425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,7168,1.7205030653211806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,7168,1.56842041015625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,6144,0.51582580142551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,65536,13.995347764756945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,6144,1.425683551364475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,5120,0.3956444528367784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,6144,1.3428915871514215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,5120,1.1356240378485787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,4096,0.31402932273017037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,5120,1.1065653695000541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,3584,0.8209928936428494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,4096,0.9004426532321506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,4096,0.9693519804212781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,3072,0.6256364186604818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,3584,0.7840160263909234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,3072,0.25653688112894696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,2560,0.2268764442867703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,2560,0.6667982207404243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,3072,0.6810311211480035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,2048,0.4563048680623372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,2048,0.21734400590260824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,3584,0.2823013464609782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,2048,0.47795645395914715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,1536,0.1502631107966105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,1024,0.23543288972642687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,768,0.18510844972398546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,1024,0.1199573278427124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,1536,0.3807342317369249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,1024,0.27846044964260525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,512,0.13216888904571533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,768,0.10753866699006821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,256,0.09663199716144139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,768,0.22824533780415854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,512,0.09646755456924438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,128,0.07733244366115995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,2560,0.5977315372890896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,512,0.18491378095414904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,256,0.0850364433394538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,1536,0.3141271008385552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,64,0.06991644700368245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,256,0.16713510619269478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,32,0.07407910956276788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,128,0.08174933327568902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,128,0.15826488865746394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,64,0.16418222586313883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,32,0.1629342238108317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,51200,18.334889729817707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,65536,1.8760159810384114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,51200,1.4650097952948675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,16384,0.38570488823784727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,51200,3.461473676893446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,16384,1.4509911007351344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,12288,0.8536755773756238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,65536,21.861987643771702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,16384,1.3151351081000435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,65536,4.499331580268012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,12288,0.3037155469258626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,10240,0.2461520036061605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,10240,0.7633831236097547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,12288,0.9432355033026801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,10240,0.7771289083692762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,8192,0.6640924347771539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,7168,0.5294684304131402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,8192,0.6481715308295356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,7168,0.1798257827758789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,6144,0.15465244981977674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,7168,0.5202924410502116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,6144,0.4861431121826172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,5120,0.32968444294399685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,51200,5.6243786282009545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,5120,0.13152533107333714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,6144,0.488225777943929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,4096,0.11876444021860759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,8192,0.20655200216505262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,5120,0.37343822585211855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,3584,0.24698310428195527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,4096,0.31282933553059894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,3584,0.0985048876868354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,3072,0.21624355845981172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,3072,0.08919111225340103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,3584,0.26007021798027885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,2560,0.19537688626183403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,3072,0.22324000464545357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,2560,0.07725510994593303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,2048,0.14781867133246526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,65536,6.979167938232422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,2048,0.06565688716040717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,1536,0.12286489539676243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,1024,0.07984889215893216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,1536,0.05434044533305698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,2048,0.15599911742740208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,768,0.06225599845250448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,1024,0.04483911063936022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,1536,0.12389955255720352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,512,0.04986844460169474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,256,0.03604977660708957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,1024,0.09437777598698933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,768,0.07695466942257352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,4096,0.3309217823876275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,128,0.02994222111172146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,256,0.03360889024204678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,64,0.025664000047577754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,128,0.033654222885767617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,256,0.06279644701215956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,32,0.0236435549126731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,128,0.05505688985188802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,64,0.05621777640448677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,32,0.05645689037111071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,2560,0.19219288561079237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,768,0.041166222757763334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,512,0.06453066402011447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,16384,0.8381244341532389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,65536,1.3412453333536785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,512,0.03799555699030558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,16384,0.36924176745944554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,12288,0.680757310655382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,51200,2.7840186225043406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,16384,1.0145031611124675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,65536,3.2844399346245656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,51200,1.0330631468031142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,12288,0.22961155573527017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,10240,0.18578578366173637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,10240,0.5663430955674913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,12288,0.6815955373975965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,8192,0.454786671532525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,8192,0.1514728864034017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,10240,0.5804897944132487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,7168,0.3917422294616699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,8192,0.4955129093594021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,7168,0.1537191073099772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,6144,0.3046986791822645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,7168,0.4069937864939372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,6144,0.12653866079118517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,51200,4.094180636935763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,5120,0.28049688869052464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,6144,0.3401057720184326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,5120,0.10116889079411824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,4096,0.19899822605980766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,4096,0.08732266558541192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,3584,0.21140266789330375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,5120,0.31192978223164874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,3072,0.1580231057273017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,3584,0.07655821906195746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,4096,0.23378043704562715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,2560,0.12632177935706243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,3072,0.07989333073298137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,3584,0.19827289051479766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,2048,0.12027200063069661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,2560,0.060808002948760986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,3072,0.17700089348687065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,1536,0.08971199724409316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,2560,0.1469848950703939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,1024,0.06558044751485188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,2048,0.1348293357425266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,1536,0.04263910982343885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,768,0.04471288786994087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,1536,0.09669511185752021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,1024,0.03698044353061252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,512,0.034252444903055825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,1024,0.0735351112153795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,768,0.03410844339264764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,256,0.025571554899215698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,65536,6.086261325412327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,512,0.030982222821977403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,768,0.059113780657450356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,128,0.026363554928037856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,512,0.051260444853040904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,256,0.027474666635195415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,32,0.01824977828396691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,128,0.02719733284579383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,128,0.042745778958002724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,64,0.04367466767628988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,32,0.044012443886862866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,2048,0.05131466521157158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,64,0.022320000661744013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,256,0.046013332075542875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,65536,1.2039502461751301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,16384,0.7822497685750326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,16384,0.30400000678168404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,51200,0.9056417677137586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,16384,0.7999253273010254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,51200,2.257131576538086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,12288,0.2132577763663398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,12288,0.634859561920166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,65536,3.0081138610839844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,10240,0.48883819580078125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,10240,0.16642488373650446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,12288,0.612061341603597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,8192,0.14314044846428767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,8192,0.42859734429253477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,10240,0.49846665064493817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,7168,0.3657626575893826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,7168,0.11777510907914902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,6144,0.32471201154920787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,8192,0.41121776898701984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,7168,0.3376311196221246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,6144,0.1381688912709554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,5120,0.24835467338562012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,6144,0.28498488002353245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,5120,0.08888533380296494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,4096,0.17431467109256318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,51200,3.561509450276693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,3584,0.1516817808151245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,3072,0.15031288729773626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,5120,0.3660506672329373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,4096,0.23698133892483184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,3584,0.1678168906105889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,2560,0.1270702282587687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,3072,0.06276444594065349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,3072,0.147944000032213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,65536,4.49342303805881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,2560,0.051860445075564914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,1536,0.0771155556042989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,2048,0.04488355583614773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,2048,0.10495555400848389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,1024,0.053786665201187134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,1536,0.040523555543687605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,1536,0.08179733488294813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,768,0.04314488834804959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,1024,0.03126577867401971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,512,0.03308355477121141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,768,0.03022222386466132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,768,0.049711998966005116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,4096,0.08643822537528144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,256,0.022275555464956496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,3584,0.0654239985677931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,512,0.042635556724336415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,128,0.016743110285864938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,256,0.0236453331179089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,64,0.014543111125628153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,256,0.03766399953100417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,128,0.023482667075263128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,2048,0.10421066813998753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,32,0.01479111115137736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,128,0.035813334915373064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,2560,0.13855822881062826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,64,0.03576888971858554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,32,0.03614933292071024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,1024,0.06341777907477485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,512,0.027215111586782668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,65536,0.9267661836412219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,51200,0.718473752339681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,16384,0.6146062215169271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,16384,0.6524293157789443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,16384,0.26604533195495605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,12288,0.623636457655165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,12288,0.5097457567850748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,51200,1.9782630072699652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,10240,0.13459022839864096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,10240,0.49013423919677734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,65536,2.662630293104384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,10240,0.39574045605129665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,8192,0.33747198846605087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,8192,0.11571644412146674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,7168,0.2745804521772596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,8192,0.38822134335835773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,7168,0.09543199671639337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,6144,0.21868711047702363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,7168,0.2965368959638807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,12288,0.15684978167215982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,6144,0.08380444182289971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,5120,0.19645599524180093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,65536,3.6223059760199656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,6144,0.24185334311591256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,5120,0.07252622312969632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,4096,0.16757244533962676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,3584,0.1388604508505927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,4096,0.06608621941672431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,5120,0.19858488771650526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,3584,0.05431822273466322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,4096,0.16967556211683485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,3072,0.12210844622717963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,2560,0.09346489111582439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,3584,0.1364524496926202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,2560,0.04344533218277825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,3072,0.12899022632175022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,1536,0.061055110560523145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,2048,0.03672444489267137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,2560,0.10668000247743394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,2048,0.08571733368767633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,1024,0.04229688975546095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,51200,2.95356453789605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,768,0.03143466512362162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,1536,0.031162665949927434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,512,0.025200888514518738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,1024,0.025255110528733995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,1536,0.06877066691716512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,768,0.023173333870040044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,1024,0.05140355560514662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,256,0.018218666315078735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,128,0.01407999959256914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,768,0.041971554358800255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,512,0.02100444502300686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,64,0.011775111158688864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,256,0.0188719994492001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,512,0.034063110748926796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,32,0.01221777747074763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,128,0.018932445181740653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,256,0.029438220792346533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,128,0.02847644355561998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,64,0.02849066588613722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,32,0.028279999891916912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,2048,0.07464710871378581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,16384,0.434295998679267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,65536,0.9043386247422961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,51200,0.7032595740424262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,16384,0.25150754716661244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,3072,0.04761955473158094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,16384,0.5990871323479546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,12288,0.4546515676710341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,51200,1.7886106703016493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,12288,0.14696978198157415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,10240,0.2879555490281847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,65536,2.1768976847330728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,10240,0.14755377504560682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,8192,0.25798845291137695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,8192,0.11480800310770671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,10240,0.48638402091132266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,12288,0.5090524355570475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,7168,0.26029245058695477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,8192,0.28744710816277397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,7168,0.09153155485788982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,6144,0.22369421852959526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,6144,0.0952862236234877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,5120,0.16267555289798313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,7168,0.25055821736653644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,51200,2.6545751359727645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,5120,0.07247110870149401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,4096,0.0565884444448683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,65536,3.0594925350613065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,5120,0.18231377336714003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,4096,0.13907021946377224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,3072,0.10054222080442642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,3584,0.05010133319430881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,2560,0.08329333199395074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,3584,0.1277608871459961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,3072,0.04467288984192742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,2048,0.07009066475762261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,3072,0.10816978083716498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,2560,0.043419556485282056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,1536,0.055665777789221875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,2560,0.09273511171340942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,2048,0.033618665403789945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,1024,0.036792887581719294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,2048,0.07640355825424194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,1536,0.027899556689792212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,768,0.0270462218258116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,1536,0.061808890766567655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,1024,0.023616888456874426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,1024,0.04649600055482653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,512,0.020968889196713764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,768,0.02144711050722334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,256,0.01592177814907498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,768,0.03786311215824551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,4096,0.14069778389400905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,128,0.01347644461525811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,3584,0.12519200642903647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,512,0.030191110240088567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,64,0.011734222372372946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,128,0.01744622157679664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,256,0.026547556122144062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,32,0.011698666546079846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,128,0.025314667158656653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,64,0.025614221890767414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,32,0.025080000360806782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,512,0.018941332896550495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,6144,0.20804977416992188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,256,0.017451556192504037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,65536,0.7688790957132975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,16384,0.4129315482245551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,16384,0.2164177762137519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,51200,0.6011670960320367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,65536,1.8322363959418402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,12288,0.3196968767378065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,16384,0.4971822102864583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,51200,1.6570924123128254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,12288,0.13592799504597983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,10240,0.2975777784983317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,10240,0.10819733142852783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,8192,0.24807021352979872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,12288,0.4572124481201172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,7168,0.22066756089528403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,10240,0.3082746664683024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,8192,0.09734844499164158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,7168,0.07761155234442817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,6144,0.17412977748446992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,8192,0.35995644993252224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,7168,0.24173333909776476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,6144,0.18124711513519287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,51200,2.132478290134006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,5120,0.06723822487725152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,5120,0.19398310449388292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,3584,0.09972000122070312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,4096,0.050442665815353394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,3072,0.08139466577106051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,4096,0.1212044424480862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,3584,0.044171555174721606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,2560,0.07058666812049018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,3584,0.118541333410475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,3072,0.038486222426096596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,2048,0.060376001728905566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,2560,0.035022222333484225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,2560,0.08106488651699491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,1536,0.04327022367053562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,6144,0.0712311135398017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,2048,0.0668026672469245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,5120,0.14798755115932888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,65536,2.7828765445285373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,1536,0.025063110722435847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,4096,0.12291733423868816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,1536,0.05381688806745741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,768,0.023153778579499986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,512,0.017080000705189176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,1024,0.020830222302012973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,1024,0.040820442967944674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,768,0.018908444378111098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,256,0.012967999610635968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,768,0.03348533312479655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,512,0.017248888810475666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,512,0.027021333575248718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,128,0.010796444283591377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,256,0.015785778562227886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,64,0.009476444787449306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,32,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,128,0.01607733302646213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,3072,0.09508799844317967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,128,0.022368000613318548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,64,0.0226248883538776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,32,0.02234399980968899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,2048,0.029716445340050593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,1024,0.02908088763554891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,16384,0.3349164326985677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,65536,0.6422799958123101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,51200,0.5099289152357314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,256,0.02330222229162852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,16384,0.15008800559573704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,12288,0.31044888496398926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,51200,1.31806394788954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,12288,0.11976444721221924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,65536,1.6366826163397894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,10240,0.27665689256456166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,12288,0.3065546618567573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,10240,0.09800533453623454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,8192,0.19917955663469103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,8192,0.07604177792867024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,51200,1.703722635904948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,10240,0.35794133610195583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,7168,0.1918177737130059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,8192,0.3089795642428928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,6144,0.16176977422502306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,7168,0.07505066527260675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,65536,2.2563663058810763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,7168,0.17983555793762207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,5120,0.13746578163570827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,6144,0.08311555782953899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,4096,0.08826488918728298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,5120,0.05597777499092949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,3584,0.07405866516960992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,5120,0.12965777185228136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,4096,0.06126488579644097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,3072,0.06980711221694946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,4096,0.12130133310953777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,16384,0.4280693266126845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,3584,0.038586666186650596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,2560,0.05816977553897434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,3584,0.09453688727484809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,2048,0.046661333905326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,3072,0.03342755635579427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,3072,0.08074400160047743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,1536,0.04018400112787882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,2560,0.02987733483314514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,1024,0.025394666526052687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,2560,0.06941244337293836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,1536,0.02201333310869005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,768,0.021692444880803425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,1536,0.046078221665488354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,1024,0.018571555614471436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,512,0.015777778294351365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,1024,0.035291555855009295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,768,0.01682222220632765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,256,0.011402666568756104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,768,0.02905244297451443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,512,0.015765332513385348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,128,0.009398221969604492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,512,0.023510222633679707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,256,0.014715555641386243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,64,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,256,0.020304888486862183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,128,0.014477334088749357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,32,0.008628444539176093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,128,0.019363555643293593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,64,0.01958577831586202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,32,0.01955911186006334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,2048,0.026932444837358262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,2048,0.05911022424697876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,16384,0.2520684401194255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,6144,0.1567573282453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,65536,0.5388711293538412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,51200,0.4179866578843858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,51200,0.7932391166687012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,16384,0.12044355604383682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,65536,1.3794480429755316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,16384,0.348673767513699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,12288,0.10660800006654526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,10240,0.20080355803171793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,12288,0.2805528905656603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,8192,0.15169866879781088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,8192,0.06509688827726576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,7168,0.12330578433142768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,10240,0.3182968828413221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,65536,1.7626995510525172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,8192,0.16739111476474336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,7168,0.0618471105893453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,6144,0.1438479953342014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,6144,0.049406223826938205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,5120,0.09891288810306126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,7168,0.15353511439429388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,4096,0.06952000326580472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,6144,0.13629866970909968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,5120,0.041908442974090576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,3584,0.06486399968465169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,5120,0.11522577868567573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,12288,0.24406843715243867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,4096,0.036151111125946045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,3072,0.0598942240079244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,4096,0.08799199925528632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,2560,0.046738667620552905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,3584,0.03231111168861389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,2048,0.037661333878835045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,3072,0.02937600016593933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,2560,0.025813332862324182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,10240,0.07925866709815131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,1536,0.030398223135206435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,2560,0.05908888578414917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,2048,0.02198933396074507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,1024,0.020833777056799993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,2048,0.049508445792728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,1536,0.018808888064490426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,768,0.017059556312031217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,51200,1.3783742056952581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,1024,0.015959110524919298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,512,0.013068444199032254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,1536,0.039282666312323675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,1024,0.02955733405219184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,768,0.015108444624476962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,512,0.014045332868893942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,768,0.024832000335057575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,128,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,512,0.020022221737437777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,256,0.013066666821638743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,64,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,256,0.01719999975628323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,128,0.012967111335860359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,32,0.006945778098371293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,128,0.01681777834892273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,64,0.01681866745154063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,3584,0.07709066735373603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,32,0.016527111331621807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,3072,0.06778133577770658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,256,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,65536,0.4471822314792209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,65536,0.8811342451307508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,51200,0.3639022244347467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,16384,0.2599066628350152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,51200,0.8809928894042969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,16384,0.12001332971784805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,12288,0.2034426662656996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,16384,0.3324800067477756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,12288,0.1149395571814643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,51200,1.2080524232652452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,12288,0.2252462175157335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,65536,1.5152986314561632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,10240,0.07299555672539605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,8192,0.17577511734432646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,7168,0.12149066395229763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,8192,0.056227558188968234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,6144,0.1125333309173584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,8192,0.17948799663119844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,6144,0.044017778502570264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,7168,0.13600444793701172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,4096,0.06577689117855497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,6144,0.11640355322096084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,5120,0.03749155667093065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,3584,0.05780888928307427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,5120,0.1027066641383701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,4096,0.03214577833811442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,3072,0.048545777797698975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,4096,0.08068178097407024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,3584,0.02863022353914049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,10240,0.1856915553410848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,3584,0.06998578045103285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,3072,0.025920889443821375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,2048,0.033366223176320396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,3072,0.06132355663511488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,2560,0.023319111929999456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,10240,0.20480622185601127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,1536,0.026297777891159058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,2560,0.052579555246565074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,1024,0.019075555933846366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,7168,0.07223911417855157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,2048,0.04436444573932224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,768,0.015058666467666626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,1536,0.03576622075504727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,5120,0.09219377570682102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,512,0.011241777903503843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,1024,0.01551466683546702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,1024,0.026657778355810378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,256,0.00850222259759903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,768,0.014480888843536377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,768,0.0226248883538776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,512,0.013155555559529198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,128,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,512,0.018144889010323417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,64,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,256,0.01571466690964169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,32,0.006504888749784893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,128,0.012390221986505719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,128,0.015455111861228943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,2560,0.04340711236000061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,64,0.01498222185505761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,32,0.015503111812803479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,2048,0.020299555526839364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,1536,0.017836444907718234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,65536,0.4139280054304335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,16384,0.2186959981918335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,65536,0.8839991357591418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,51200,0.32576711972554523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,51200,0.8128417862786187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,256,0.012409778104888068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,12288,0.18284355269538033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,16384,0.09774666362338597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,16384,0.27798665894402397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,10240,0.17007554901970756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,12288,0.08994577990637885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,8192,0.13437422116597494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,10240,0.06456355253855388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,51200,0.9862702687581381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,12288,0.2135875489976671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,7168,0.12188266383277045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,10240,0.188993771870931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,7168,0.0471111105548011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,8192,0.13404177294837102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,5120,0.0803297758102417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,6144,0.039982222848468356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,6144,0.11796532736884223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,5120,0.03425688876046075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,4096,0.06443733639187284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,5120,0.08790133396784465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,3584,0.05376977721850077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,4096,0.02926844358444214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,4096,0.07141333156161837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,3072,0.04649244414435493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,3584,0.06221955352359348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,65536,1.2548222011990018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,3072,0.02335022224320306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,2560,0.039649777942233615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,3072,0.054917332198884755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,8192,0.06339999702241686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,2048,0.03352800011634827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,2560,0.020968889196713764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,1536,0.023194667365815904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,2048,0.01849333279662662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,7168,0.11824622419145371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,1024,0.016799110505315993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,2048,0.039741333987977766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,1536,0.016241777274343703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,768,0.013097777962684631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,1024,0.014108444253603617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,1536,0.03176888823509216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,512,0.010799110763602786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,1024,0.024114666713608637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,768,0.01309511148267322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,768,0.020432000358899433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,512,0.012325333224402534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,3584,0.02640266716480255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,128,0.007248000138335758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,256,0.011905777785513135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,256,0.014204444156752693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,64,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,32,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,128,0.013815999858909182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,6144,0.09644711017608643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,64,0.013452444639470843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,32,0.013789333403110504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,2560,0.047148442930645414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,65536,0.3607689009772406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,256,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,51200,0.5318453576829698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,65536,0.8406337632073296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,512,0.01651733285850949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,16384,0.17416977882385254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,51200,0.2842515574561225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,128,0.011552000211344825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,12288,0.15935911072625056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,51200,0.7794737815856934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,65536,1.0179164674546983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,16384,0.24644356303744844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,10240,0.12775111198425293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,12288,0.07694222529729207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,8192,0.09722044732835557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,12288,0.17463110552893746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,10240,0.06233955754174126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,7168,0.08961955706278484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,8192,0.04684177703327603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,6144,0.08063111040327284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,10240,0.14574044280582005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,7168,0.04181066817707486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,8192,0.12000621689690484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,5120,0.0641804469956292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,6144,0.03645066751374139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,4096,0.046317332320743136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,7168,0.11342755953470866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,3584,0.03944622145758735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,6144,0.09361422061920166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,3072,0.036834667126337685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,4096,0.02684088879161411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,5120,0.076328886879815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,3584,0.0247688889503479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,4096,0.06360799736446805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,2560,0.030560887522167627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,2048,0.024912888805071514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,3584,0.05588889122009277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,3072,0.02197422252760993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,1536,0.01997422178586324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,2560,0.020853334003024634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,3072,0.04875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,2048,0.017587555779351127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,1024,0.014555555250909595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,2560,0.04209511147605049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,16384,0.11332533094618057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,768,0.011401777466138205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,2048,0.03561333484119839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,512,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,1024,0.013279111021094851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,1536,0.02830933199988471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,256,0.006524444454246097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,1024,0.021465778350830078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,512,0.011699555648697747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,768,0.018236445056067575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,128,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,512,0.014825777875052558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,256,0.01106844428512785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,64,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,256,0.01275022245115704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,32,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,128,0.012263110942310758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,5120,0.03188000122706095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,64,0.012116444607575735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,32,0.012448000411192576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,65536,0.30326665772332084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,65536,0.5972640249464247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,1536,0.01536355581548479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,51200,0.4426000118255615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,768,0.012763555679056378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,51200,0.2661822107103136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,16384,0.07646133502324422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,16384,0.20762577321794298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,12288,0.11685421731736924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,10240,0.09453244341744317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,12288,0.06329955657323201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,51200,0.6722168922424316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,12288,0.14983022212982178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,8192,0.07861777808931139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,10240,0.05274844500753614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,7168,0.06523822413550483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,8192,0.04092355569203695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,10240,0.13435288270314535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,6144,0.05093510945638021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,8192,0.10569510857264201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,7168,0.03568089008331299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,5120,0.047041777107450694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,7168,0.08919733100467259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,6144,0.03199199835459391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,4096,0.03761333227157593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,16384,0.1603955560260349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,5120,0.028800888193978205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,3584,0.035226666265063815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,5120,0.06492977672153048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,4096,0.023745778534147475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,3072,0.029135111305448744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,3584,0.021531555387708876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,4096,0.053252445326911084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,128,0.010955555571450127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,2560,0.024940444363488093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,3584,0.04778933193948534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,2048,0.02066044509410858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,3072,0.019670221540662978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,1536,0.016749334004190233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,2560,0.017921777235137094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,2048,0.016151999433835346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,1024,0.012437333663304647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,2048,0.031184001101387873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,1536,0.014314666390419006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,768,0.010261333651012844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,1536,0.024710221423043147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,1024,0.012498666842778524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,1024,0.018956444329685636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,512,0.008420444197124904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,768,0.011891555454995898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,256,0.007019555403126611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,512,0.01088444474670622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,6144,0.08024533589680989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,128,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,256,0.010382222632567087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,256,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,64,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,65536,0.8303866916232638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,128,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,32,0.004871110949251386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,128,0.010750222537252637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,64,0.0107360002067354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,3072,0.042318221595552236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,32,0.01089422239197625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,2560,0.036887112591001726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,65536,0.4070151117112901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,16384,0.11835732724931504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,65536,0.2497182157304552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,51200,0.36963555547926163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,51200,0.2016702228122287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,768,0.01589777734544542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,16384,0.07173511054780748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,65536,0.6670239766438802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,512,0.013112000293201871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,10240,0.08620888657040066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,51200,0.5213200251261393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,16384,0.1674097776412964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,12288,0.12495377328660752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,8192,0.0648497806655036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,7168,0.05592088566886055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,10240,0.04522755410936144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,8192,0.034640000926123724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,6144,0.04455022348297966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,10240,0.10596088568369548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,8192,0.08469866381751166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,7168,0.030735111898846094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,5120,0.037315554089016385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,7168,0.07555111249287923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,4096,0.02830044428507487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,6144,0.02732533382044898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,5120,0.023952888117896184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,3584,0.025655110677083332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,4096,0.02033155494266086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,5120,0.05457422137260437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,3072,0.022210665875011023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,3584,0.018631110588709515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,3584,0.04091644287109375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,2560,0.019425777925385367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,3072,0.03589955634540982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,2560,0.015860444969601102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,12288,0.052684442864524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,2048,0.016072000066439312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,2560,0.03089955449104309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,1536,0.012712000144852532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,2048,0.014535999960369535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,2048,0.025941333836979334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,1024,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,1536,0.012793777717484368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,1536,0.0214684440029992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,768,0.007605333295133378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,6144,0.06510666343900892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,1024,0.01127822200457255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,512,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,1024,0.016210666961140104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,768,0.014071111877759298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,4096,0.04514222343762716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,256,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,512,0.010038221875826517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,128,0.004573333180612988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,512,0.011735111474990845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,3072,0.017070222232076857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,12288,0.09071911043590969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,64,0.0042160000238153665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,256,0.010006222460005019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,128,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,128,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,64,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,32,0.009306666751702627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,65536,0.29115645090738934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,65536,0.19681689474317762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,51200,0.2282808886633979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,16384,0.09028889073265924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,65536,0.5207840071784126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,51200,0.15686578220791286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,768,0.01069777790043089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,51200,0.4752088652716742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,12288,0.06234044498867459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,16384,0.053364445765813194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,10240,0.05161866545677185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,256,0.009718221922715506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,16384,0.13238399558597141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,8192,0.043185777134365506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,12288,0.04307288924853007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,7168,0.036757333411110774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,10240,0.038035555018319026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,12288,0.10060089164310032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,8192,0.029578665892283123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,10240,0.08429511388142903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,5120,0.024708444873491924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,8192,0.0682693322499593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,7168,0.060975107881757945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,4096,0.01995466649532318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,6144,0.05236444539493985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,5120,0.019486222002241347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,3584,0.01754577789041731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,5120,0.044347554445266724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,4096,0.017098666893111337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,3072,0.015864888827006023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,4096,0.03710844450526767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,3584,0.01573511130279965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,2560,0.013389333254761167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,3584,0.03358755509058634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,3072,0.01458755632241567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,2048,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,3072,0.029570665624406602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,2560,0.013382222089502545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,1536,0.009342222577995723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,2048,0.012228444218635559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,2048,0.021718222233984206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,32,0.004212444441186057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,1024,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,1536,0.01112711098459032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,768,0.006335111127959357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,1536,0.017889777819315594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,7168,0.02563200063175625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,1024,0.010272000398900773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,512,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,6144,0.022579555710156757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,256,0.004254222330119875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,1024,0.013759111364682516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,512,0.009379555781682333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,768,0.012016889121797351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,128,0.003539555602603488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,512,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,256,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,64,0.0032186667538351486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,32,0.003532444437344869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,256,0.008627555436558193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,128,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,128,0.00795377790927887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,64,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,32,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,51200,0.20391555627187094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,65536,0.19355555375417074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,16384,0.06310044394599067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,65536,0.43874666425916886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,6144,0.031129777431488037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,51200,0.1656693352593316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,51200,0.3450373278723822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,12288,0.049399111005995006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,16384,0.05295999844868978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,768,0.00999022192425198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,10240,0.04130666785769992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,16384,0.1137475570042928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,12288,0.04412177867359585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,8192,0.03223199976815118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,12288,0.08596444129943848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,10240,0.03646488984425863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,7168,0.02697511182890998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,10240,0.0721528861257765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,8192,0.027240888939963445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,6144,0.022775999373859827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,8192,0.05878844526078966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,5120,0.01959022217326694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,65536,0.23399911986456978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,7168,0.0517137779129876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,6144,0.02179644505182902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,4096,0.016018667154841952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,6144,0.045200887653562755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,2560,0.025454221500290766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,5120,0.019291554888089497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,3584,0.014525334040323893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,3072,0.012472888661755456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,5120,0.038273778226640486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,4096,0.03195110956827799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,3584,0.015643555257055495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,2560,0.011057777537239922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,3584,0.028254220883051556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,2048,0.009023110899660323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,3072,0.02552266584502326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,2560,0.013455111119482251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,2560,0.022061333060264587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,1536,0.007770666645632849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,1024,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,2048,0.012272888587580787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,1536,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,768,0.005500444521506627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,1024,0.010203555226325989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,1024,0.01223288890388277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,7168,0.024361777636739943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,768,0.00998488896422916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,768,0.010803555448849997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,256,0.003565333369705412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,512,0.009022222624884712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,512,0.009186666872766282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,128,0.003170666595300039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,256,0.008770667016506195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,256,0.007795555724038019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,64,0.003185777821474605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,128,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,3072,0.014513777361975776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,32,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,64,0.007111111448870764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,65536,0.21498933103349474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,2048,0.018945778409639995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,1536,0.015790222419632804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,51200,0.16864355405171713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,512,0.004418666578001446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,65536,0.14709422323438856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,16384,0.056764443715413414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,65536,0.3863191074795193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,51200,0.11788177490234375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,16384,0.041280001401901245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,4096,0.016827555166350473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,12288,0.04302933481004503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,51200,0.30209599600897896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,16384,0.10169511371188694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,10240,0.03465422325664096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,8192,0.026799110902680293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,12288,0.03272533416748047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,10240,0.029073778125974867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,7168,0.023215111758973863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,12288,0.07673244343863593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,8192,0.020733333296246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,6144,0.020234665936893888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,10240,0.06428800026575725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,8192,0.05209155546294319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,7168,0.018336888816621568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,6144,0.01667644414636824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,7168,0.04628088739183214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,4096,0.014607111612955729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,32,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,6144,0.04052711195415921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,3584,0.012899555265903473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,5120,0.034276445706685386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,3072,0.01143200033240848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,4096,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,3584,0.012400888734393649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,4096,0.0284186667866177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,2560,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,3072,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,2048,0.008377778033415476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,3072,0.0229120006163915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,2560,0.011090666883521609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,1536,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,2560,0.02016266683737437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,2048,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,2048,0.017254221770498488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,1024,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,1536,0.009705777797434065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,128,0.008784000244405534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,1536,0.014259555273585849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,1024,0.009198221895429824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,5120,0.017601778109868366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,512,0.004212444441186057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,1024,0.011178666518794166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,768,0.008802666432327693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,5120,0.01493866741657257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,768,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,256,0.0036231109665499795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,512,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,128,0.0032115555885765287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,512,0.008648888932334052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,256,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,64,0.002855111120475663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,128,0.007948444121413762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,256,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,3584,0.025602666868103877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,128,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,65536,0.12469332747989231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,32,0.00646666685740153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,51200,0.10182666778564453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,65536,0.1457164420021905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,16384,0.03796266516049703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,65536,0.31152711974249947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,51200,0.11694310771094428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,768,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,12288,0.026168000366952684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,51200,0.24488088819715711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,16384,0.04090133309364319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,10240,0.0203840004073249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,16384,0.08135466443167792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,12288,0.03251466817326016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,8192,0.016301333904266357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,12288,0.06190399991141426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,10240,0.028491556644439697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,7168,0.014843554960356818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,8192,0.0200284438000785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,10240,0.051910221576690674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,6144,0.013036444783210754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,32,0.0028533333291610083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,7168,0.018512000640233357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,5120,0.01107377807299296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,7168,0.0377475553088718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,6144,0.016639999217457242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,6144,0.033334222104814314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,64,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,3584,0.008661333057615492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,5120,0.028246220615175035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,4096,0.013224889006879596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,3072,0.007692444655630324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,4096,0.02353777819209629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,3584,0.012773333324326409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,2560,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,3584,0.02107288936773936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,3072,0.011725333001878528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,2048,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,3072,0.019123555885420907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,2560,0.011129777464601727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,1536,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,2560,0.01682844426896837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,2048,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,1024,0.004447999927732679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,1536,0.010011555420027839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,2048,0.014482667048772177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,768,0.003858666867017746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,1536,0.012154666913880242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,1024,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,512,0.0032177778581778207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,8192,0.04213866591453552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,1024,0.009735110733244155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,256,0.0028497777465316984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,512,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,768,0.008664888640244802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,4096,0.009311999711725447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,128,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,512,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,256,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,64,0.0027946666296985415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,5120,0.014908444550302295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,32,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,256,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,128,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,65536,0.08359466658698188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,128,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,51200,0.06830133332146539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,64,0.00564177789621883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,32,0.005650666852792104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,16384,0.027339556150966223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,12288,0.020577778418858845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,65536,0.14477599991692439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,51200,0.11636622746785481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,16384,0.0417795545525021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,65536,0.2683315542009142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,10240,0.017087999317381117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,12288,0.03321866525544061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,51200,0.2104426622390747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,8192,0.012759110993809171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,12288,0.05372622278001574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,10240,0.028552889823913574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,7168,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,10240,0.04536622100406223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,6144,0.010209778116808997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,7168,0.019541333119074505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,8192,0.03685422076119317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,768,0.00868977771864997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,5120,0.00909511082702213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,7168,0.03289688958062066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,4096,0.007255111303594377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,6144,0.017860444055663217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,5120,0.01587022178702884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,3584,0.006657777975002925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,4096,0.013789333403110504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,5120,0.024725332856178284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,3072,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,4096,0.02081155611409081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,3584,0.013109332985348172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,2560,0.005407111098368962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,3584,0.018961777289708454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,3072,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,2048,0.004632888982693355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,3072,0.016970665918456185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,2560,0.01142133358452055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,1536,0.0041982221106688184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,16384,0.070342222849528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,2560,0.015150222513410779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,2048,0.010678222609890832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,1024,0.0035013332962989807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,2048,0.013155555559529198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,768,0.003149333306484752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,1536,0.010058666268984476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,8192,0.020976000361972388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,512,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,1536,0.011145778000354767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,1024,0.008999110923873054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,768,0.008711111214425828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,256,0.0024622221373849446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,512,0.008307555483447181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,128,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,512,0.00700711127784517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,6144,0.028831998507181805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,256,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,64,0.0031146665828095544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,32,0.0021475555582178966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,256,0.005803555664088991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,128,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,128,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,65536,0.07828711138831244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,64,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,51200,0.06311821937561035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,32,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,16384,0.023172444767422144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,12288,0.0138568886452251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,10240,0.012228444218635559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,8192,0.011380444798204633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,65536,0.2679164409637451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,51200,0.2102737824122111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,16384,0.07022133138444689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,12288,0.05382844474580553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,6144,0.009570666485362584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,5120,0.008227555288208855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,8192,0.03672177924050225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,1024,0.009203555683294931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,4096,0.0069208890199661255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,3584,0.006319111006127463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,7168,0.03249511122703552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,3072,0.005884444548024072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,6144,0.028393778536054824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,768,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,2560,0.005186666631036335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,2048,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,4096,0.020824000239372253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,1536,0.0038862224254343244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,3584,0.018810666269726224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,1024,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,3072,0.01700177788734436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,2560,0.01516711049609714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,2048,0.013239111337396832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,512,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,1536,0.011056000159846412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,256,0.002456888970401552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,1024,0.008976000050703684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,128,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,512,0.006991110742092133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,64,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,10240,0.04523644513554043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,128,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,32,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,65536,0.07363200187683105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,64,0.0053013331360287136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,51200,0.05895911322699653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,5120,0.02474400069978502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,16384,0.020761777957280476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,768,0.0028808888875775864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,65536,0.26852620972527397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,12288,0.012507555385430654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,51200,0.21006400055355498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,16384,0.07017866770426433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,10240,0.010667555862002902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,8192,0.009335111412737105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,768,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,7168,0.009072889056470659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,12288,0.05350133445527819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,6144,0.008459555606047312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,10240,0.04527733392185635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,5120,0.010557333628336588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,8192,0.03670310974121094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,4096,0.008794666992293464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,3584,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,7168,0.03250311149491204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,3072,0.007233777807818518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,6144,0.02860266632503933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,5120,0.024495111571417913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,2560,0.00638755535085996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,4096,0.02086488902568817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,32,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,3584,0.01887999971707662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,1536,0.003858666867017746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,1024,0.003167111012670729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,3072,0.01683822274208069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,768,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,2560,0.01517599986659156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,512,0.0026897777699761917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,2048,0.013381332986884646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,1536,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,1024,0.008987555901209513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,128,0.0021226666867733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,768,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,64,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,512,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,32,0.001806222316291597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,256,0.00574933323595259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,128,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,7168,0.01068800025516086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,64,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,32,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,2048,0.004598222259018156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,256,0.002395555583967103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,16384,1.4741155836317275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,16384,4.352793799506293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,16384,5.085210588243273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,12288,3.706116570366753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,51200,5.843388451470269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,16384,4.5076984829372835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,12288,1.068895975748698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,12288,3.498641755845812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,10240,0.8788151211208768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,12288,3.243140538533529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,10240,2.8948088751898875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,10240,3.498977873060438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,8192,0.7149279912312826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,8192,2.3869654337565103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,10240,2.7118737962510853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,8192,2.4524612426757812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,8192,2.204837375217014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,51200,13.892668830023872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,7168,0.6184808943006728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,7168,2.224600897894965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,7168,1.9372738732231989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,6144,0.547651555803087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,7168,1.9694133334689672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,6144,1.6838809119330511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,5120,1.38465330335829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,5120,0.47704267501831055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,6144,1.662236425611708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,6144,1.6783857345581055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,5120,1.4698301951090496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,4096,1.0930755403306749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,5120,1.287718243069119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,4096,0.4127946694691976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,4096,1.1573378245035808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,3584,0.9314346313476562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,4096,1.0367973115709093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,3584,0.35233600934346515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,3072,0.3053750991821289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,3584,0.970553821987576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,3072,0.7907013363308377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,3584,0.9391351275973849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,3072,0.8663422266642252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,3072,0.8181733555263944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,2560,0.27713510725233287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,2560,0.7051804330613879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,2048,0.5557831128438314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,2048,0.23222756385803223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,2560,0.6702159775627984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,2048,0.6282906532287598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,2048,0.5463155640496148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,1536,0.18434222539265951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,2560,0.8132568995157877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,1536,0.45733600192599827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,1024,0.30261511272854275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,1536,0.4450097613864475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,1024,0.14822044637468126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,1024,0.35236088434855145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,1024,0.31438133451673717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,768,0.2762355539533827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,1536,0.4614728821648492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,512,0.15798755486806235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,768,0.2589493327670627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,768,0.12724710835350886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,256,0.11789777543809678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,512,0.11192177401648627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,768,0.2646639876895481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,512,0.20907110638088652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,256,0.2109982172648112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,128,0.09654400083753799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,256,0.1640062199698554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,51200,22.445741441514755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,128,0.19815644952985975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,64,0.09409422344631618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,128,0.12519733111063638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,32,0.09441866477330525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,512,0.23085689544677734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,64,0.2007368935479058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,32,0.20347021685706246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,256,0.09832355711195205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,128,0.0968773365020752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,65536,5.704040951199001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,51200,10.709469265407987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,51200,4.2734658983018665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,65536,13.963835822211372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,16384,3.437774234347873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,16384,1.114918178982205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,51200,12.366019354926216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,16384,4.353123558892144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,51200,16.788304646809895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,12288,0.8652008904351128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,16384,3.5749342176649304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,12288,2.686808056301541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,12288,2.7434452904595266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,65536,15.172339545355904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,12288,2.6055545806884766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,10240,2.488835652669271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,10240,0.683659553527832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,10240,2.3995653788248696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,65536,21.86404758029514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,8192,0.5866595374213325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,10240,2.1103138393825955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,8192,1.9612774319118922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,7168,1.6029528511895075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,7168,0.5000275505913628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,7168,1.6366764704386394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,8192,1.6293200386895075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,6144,1.3590301937527125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,6144,0.43758222791883683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,8192,1.8859912024603949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,7168,1.4170444276597765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,6144,1.3203190697564018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,5120,1.2309128443400066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,5120,0.382060448328654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,6144,1.22815916273329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,4096,0.3119493325551351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,5120,1.1555573145548503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,4096,0.8780897988213433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,4096,0.8913475672403971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,5120,1.0728435516357422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,3584,0.8853537771436902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,3584,0.29804354243808323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,4096,0.8199831114874946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,3072,0.6060862011379665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,3072,0.24429334534539116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,3584,0.6998693148295084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,3072,0.6787964502970377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,2560,0.5770488844977485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,2560,0.21078044838375518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,3072,0.654712888929579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,2048,0.17767288949754503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,2560,0.5216106838650173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,3584,0.7864569028218588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,1536,0.3566506703694661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,2048,0.766849782731798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,2560,0.5584879981146919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,1536,0.367465787463718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,1024,0.23338045014275444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,1536,0.16477599408891466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,2048,0.3893253273434109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,1536,0.3436017831166585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,1024,0.27400265799628365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,768,0.18169777923160127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,768,0.10195999675326878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,1024,0.24958933724297416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,768,0.2056302229563395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,768,0.21355999840630424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,512,0.18268356058332655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,2048,0.42746488253275555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,256,0.09646044174830119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,512,0.08951022227605183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,128,0.07636088795132108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,256,0.07905599806043836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,256,0.16431377993689641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,1024,0.11667911211649577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,256,0.13089422384897867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,64,0.06348266866472033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,128,0.07756000094943576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,32,0.06559999783833821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,128,0.09958844714694554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,64,0.15829777717590332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,32,0.16073155403137207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,512,0.16631821791330972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,128,0.15546133783128527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,512,0.1361128886540731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,65536,1.8863387637668187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,65536,4.631664699978299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,51200,1.3933689329359267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,51200,4.128935919867621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,16384,1.2655910915798612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,16384,1.3601457807752821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,16384,0.3782009018792047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,51200,5.400864919026692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,16384,1.2145804299248588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,65536,5.034608205159505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,65536,7.431017557779948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,51200,3.8737865024142795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,12288,0.9867626825968424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,12288,0.2902408970726861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,12288,0.9593093660142687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,12288,0.8153866661919488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,10240,0.2623982164594862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,10240,0.9058408737182617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,8192,0.6948960092332629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,10240,0.7889955308702258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,8192,0.18990222613016763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,8192,0.5910293261210123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,10240,0.7042640050252279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,7168,0.6059022479587132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,8192,0.5170488887363011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,6144,0.44800355699327254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,6144,0.14721511469946966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,7168,0.7282177607218424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,6144,0.47535377078586155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,5120,0.3729697863260905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,5120,0.12946044074164495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,6144,0.43291200531853574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,4096,0.29998310407002765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,5120,0.37196532885233563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,7168,0.17167645030551484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,4096,0.10629244645436604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,3584,0.22482310401068795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,4096,0.2920337782965766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,4096,0.27345334159003365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,3584,0.09514666928185357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,3072,0.21440445052252877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,3584,0.25127379099527997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,7168,0.4530791176689996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,3072,0.08521866798400879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,2560,0.1865288946363661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,3584,0.2421493265363905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,3072,0.21984444724188912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,2048,0.1565262211693658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,2560,0.1883066627714369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,3072,0.20829066965315077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,1536,0.11143733395470513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,5120,0.3373271094428168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,2560,0.18086043993631998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,1536,0.05265955461396111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,2048,0.14991645018259683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,1024,0.07730488644705878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,1536,0.12968888547685412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,51200,15.240763346354166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,1536,0.12009244494967991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,768,0.07734133137596978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,1024,0.04417511158519321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,1024,0.0935315555996365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,512,0.05570755402247111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,768,0.0399528874291314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,1024,0.09172177977032131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,768,0.0745039979616801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,2560,0.07433778047561646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,256,0.039184000757005476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,512,0.06376800272199842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,512,0.03587822119394938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,768,0.10667733351389568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,128,0.0266977780395084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,2048,0.06297955248090956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,512,0.06386133035024007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,64,0.022513777017593384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,128,0.07034577925999959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,128,0.03182222114668952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,32,0.02754222187731001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,128,0.0384880006313324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,64,0.05492533246676127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,32,0.05549600058131748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,2048,0.15447644392649332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,256,0.05722577704323662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,256,0.05061599943372938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,65536,1.3678453233506944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,256,0.03237777617242601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,16384,0.9048728942871094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,65536,3.358538733588325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,51200,0.9778639475504557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,51200,2.8932505713568797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,16384,0.31809067726135254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,16384,0.9727741877237955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,12288,0.6694106525844998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,16384,0.86935822168986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,12288,0.7156942155626086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,12288,0.23257777425977919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,51200,3.0117270151774087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,65536,3.7765252855088978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,51200,4.377235412597656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,12288,0.6090186436971029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,10240,0.5662773450215658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,10240,0.1929697725507948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,8192,0.46713776058620876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,8192,0.14594844977060953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,65536,6.355491638183594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,8192,0.46598132451375324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,10240,0.5794711112976074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,10240,0.5072959793938531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,7168,0.4325173430972629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,7168,0.13224621613820395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,8192,0.5034329096476237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,6144,0.35712356037563747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,7168,0.39435378710428876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,6144,0.11681600411732991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,6144,0.3442995548248291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,5120,0.2829066647423638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,7168,0.3591875500149197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,5120,0.099945776992374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,4096,0.22789955139160156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,5120,0.29820089870029026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,6144,0.31284356117248535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,4096,0.08316710922453138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,3584,0.18017244338989258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,4096,0.2342017756568061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,3584,0.07407466570536296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,5120,0.256784889433119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,3072,0.16697866386837432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,3584,0.19255733489990234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,2560,0.11966221862369114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,3072,0.06604711214701335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,4096,0.21123822530110678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,3584,0.18727021747165254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,3072,0.17673067251841226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,2048,0.1025173332956102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,2560,0.05760888920889961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,1536,0.07457955678304036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,2560,0.14639111359914145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,2048,0.05067200130886502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,2048,0.1193502214219835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,1024,0.06702399916119046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,1536,0.042240000433391996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,2560,0.13940978050231934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,1536,0.09485422240363227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,768,0.050197333097457886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,2048,0.11853955851660834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,1024,0.03573244478967454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,1024,0.07253422339757283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,1536,0.09523200326495701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,512,0.04522666666242811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,768,0.032574223147498235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,768,0.05814488728841146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,256,0.0330079992612203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,1024,0.07409155368804932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,512,0.029278222057554457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,128,0.02818044357829624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,512,0.049364444282319814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,768,0.062055110931396484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,256,0.02663466665479872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,64,0.02572088936964671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,256,0.04445688923199972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,512,0.051495111650890775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,128,0.02606933315594991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,32,0.02677244444688161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,256,0.04074133435885111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,3072,0.16556799411773682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,64,0.042656888564427696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,32,0.043432888057496816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,128,0.04211911227968004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,128,0.0309840010272132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,65536,1.0853502485487196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,16384,0.8600311279296875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,51200,0.8887999852498373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,51200,2.4755564795600042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,65536,2.9978756374782987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,16384,0.262060456805759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,12288,0.5753368801540798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,16384,0.8427155812581381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,51200,3.6900507609049478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,12288,0.18137778176201713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,16384,0.7416444354587131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,65536,4.24252192179362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,65536,3.204314761691623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,10240,0.47721068064371747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,10240,0.15347999996609157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,12288,0.518687989976671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,51200,2.5104132758246527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,10240,0.4769022199842665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,8192,0.38433154424031574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,8192,0.12517333030700684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,10240,0.43123377694023984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,7168,0.3324515554640028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,8192,0.5669377644856771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,7168,0.32894576920403373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,6144,0.280403561062283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,12288,0.5840284559461806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,8192,0.3486631181504991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,7168,0.3383315669165717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,5120,0.23555911911858451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,6144,0.11480622821384007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,6144,0.31431378258599174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,5120,0.08506488800048828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,4096,0.19723910755581328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,5120,0.2436915503607856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,6144,0.2693235609266493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,3584,0.13730843861897787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,4096,0.19188800123002794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,5120,0.22953422864278158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,3584,0.06412622001436022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,3584,0.16440089543660483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,4096,0.18718577755822075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,7168,0.11147377226087783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,2560,0.11805421776241726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,3072,0.057256890667809375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,3072,0.14384978347354466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,2560,0.05023733443684048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,3072,0.1419395605723063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,2560,0.12322489420572917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,2048,0.09914666414260864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,2560,0.12140356169806586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,2048,0.043772445784674756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,2048,0.10167111290825738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,1536,0.07479200098249647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,1536,0.03691733214590285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,2048,0.10249955786599053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,1024,0.04775555597411262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,3072,0.13564356168111166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,1536,0.08145688639746772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,768,0.039889779355790876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,1024,0.030925333499908447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,1536,0.08327911297480266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,512,0.028415110376146104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,1024,0.061786664856804736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,768,0.05013066530227661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,768,0.028157333532969158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,256,0.01978488928741879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,1024,0.06454488966200086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,512,0.02537600033813053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,768,0.055043554968304105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,128,0.018945778409639995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,512,0.04199733336766561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,256,0.022651554809676275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,256,0.03753511110941569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,512,0.04576177729500664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,64,0.01587288909488254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,32,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,128,0.022088888618681166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,4096,0.07395288679334852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,256,0.03608266512552897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,128,0.027653333213594224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,64,0.03536355495452881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,32,0.03548977772394816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,3584,0.16429866684807673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,128,0.03544266687499152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,65536,0.8926844067043729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,51200,0.6779457728068033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,65536,2.492983076307509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,51200,2.285907533433702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,16384,0.19771733548906115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,16384,0.7224062283833822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,51200,3.086939493815104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,51200,2.0460177527533636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,16384,0.8068426450093588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,65536,2.6871840159098306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,12288,0.16382843918270534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,65536,4.0451100667317705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,12288,0.47277243932088214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,10240,0.3923955493503147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,12288,0.4320995542738173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,10240,0.4021626578436957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,8192,0.255928013059828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,7168,0.21719911363389757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,8192,0.1030728883213467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,8192,0.3173786799112956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,10240,0.3597982194688585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,12288,0.5306693183051215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,7168,0.09302044577068752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,6144,0.173598223262363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,16384,0.5954462157355415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,8192,0.29148533609178334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,7168,0.2751057677798801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,5120,0.17091288831498888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,6144,0.0810097787115309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,7168,0.2526640097300212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,6144,0.22694667180379233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,4096,0.1254053380754259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,5120,0.07048710849550036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,5120,0.19497333632575142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,6144,0.22092621856265596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,3584,0.12478578090667725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,4096,0.058430221345689565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,10240,0.12820443842146131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,4096,0.15259822209676108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,5120,0.18567111757066515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,3584,0.05348889032999674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,4096,0.15348888768090144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,3584,0.13573155138227674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,3072,0.11829688813951279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,2560,0.09484977854622735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,3584,0.13628355662027994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,2048,0.06995555427339342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,2560,0.04216177927123176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,3072,0.11959644158681233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,2560,0.10214755270216201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,1536,0.0565333366394043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,2048,0.03636444608370463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,2560,0.1031893359290229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,2048,0.08415200312932332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,1024,0.03903288973702325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,1536,0.06696533494525485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,1536,0.029607110553317603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,2048,0.08728888962003921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,768,0.03432000014517043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,1024,0.024118221468395654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,3072,0.09018577469719781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,1536,0.07112266620000203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,512,0.024837333295080397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,768,0.022857778602176245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,768,0.041402667760849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,3072,0.04785777793990242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,1024,0.05593066745334201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,256,0.020260444945759244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,512,0.020239111449983384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,512,0.03324177861213684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,768,0.046556442975997925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,128,0.014463111758232117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,256,0.029186666011810303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,256,0.018530666828155518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,64,0.012315555579132505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,512,0.03938311007287767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,32,0.014489778214030795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,256,0.030849777989917334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,128,0.018377777602937486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,128,0.027902222341961328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,128,0.023548444112141926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,64,0.02809777855873108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,32,0.027914666467242773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,1024,0.0506533318095737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,65536,0.9000444412231445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,51200,0.6856293148464627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,65536,2.00237062242296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,51200,1.8217705620659723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,16384,0.7111040221320258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,16384,0.20893332693311903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,16384,0.725379572974311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,12288,0.4359297752380371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,51200,2.3413217332628036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,51200,1.8293387095133464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,12288,0.15221777227189806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,16384,0.5183146794637045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,12288,0.4303075472513835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,10240,0.3382062117258708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,65536,2.325094223022461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,12288,0.3748648961385091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,10240,0.11856533421410455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,65536,3.380707634819878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,8192,0.2804364363352458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,8192,0.09374311235215928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,10240,0.3850088914235433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,7168,0.2538355456458198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,8192,0.3117528756459554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,7168,0.08345066838794285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,10240,0.3182755576239692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,6144,0.21252177821265328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,8192,0.26407289505004883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,5120,0.16708978017171225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,6144,0.21881422731611463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,7168,0.23892090055677626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,5120,0.07102666960822211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,5120,0.18246844079759386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,6144,0.19888799720340303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,3584,0.11320711506737603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,4096,0.053744001521004565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,5120,0.16552889347076416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,4096,0.13731022675832114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,3072,0.08591288990444607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,3584,0.04804977774620056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,4096,0.13614488972557917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,3584,0.12113511562347412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,2560,0.08204088608423869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,3072,0.04225600096914503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,3072,0.10564800103505452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,7168,0.29402756690979004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,6144,0.07337688737445407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,2048,0.06557777855131361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,3072,0.1067919996049669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,2560,0.09098577499389648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,1536,0.05316977699597677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,4096,0.1398586697048611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,2560,0.09242400195863511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,2048,0.07504266500473022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,1024,0.034281777011023626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,2048,0.07891111241446601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,1536,0.02646222213904063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,1536,0.05971199936336941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,1536,0.06492266390058729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,1024,0.04526399903827243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,1024,0.02223555578125848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,768,0.02741688821050856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,768,0.02030933399995168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,1024,0.050680889023674854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,3584,0.1229599979188707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,512,0.019928000039524503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,768,0.037224001354641385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,256,0.017549332645204332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,512,0.0184879998366038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,2560,0.03694399860170152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,768,0.04281599985228645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,512,0.030031998952229817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,128,0.012127999630239276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,256,0.01679733395576477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,256,0.025980444418059453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,512,0.03560533457332187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,64,0.013792888985739814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,32,0.014408888088332282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,128,0.016600888636377122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,128,0.024999999337726172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,256,0.028068443139394123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,128,0.021312000022994146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,64,0.025179555018742878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,32,0.024792888098292883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,2048,0.03160355488459269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,65536,0.7052506870693631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,51200,1.5810435612996419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,16384,0.6363004578484429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,51200,0.5793084568447536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,65536,2.197535196940104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,16384,0.17818221780988905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,51200,2.104463153415256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,16384,0.501056883070204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,12288,0.41795023282368976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,51200,1.7394719653659398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,65536,2.8012542724609375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,12288,0.12315111690097386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,16384,0.4738168716430664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,65536,2.0760951571994357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,12288,0.37319376733567977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,10240,0.290975120332506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,10240,0.10395999749501546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,8192,0.22850134637620714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,12288,0.34855466418796116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,8192,0.0938791102833218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,10240,0.369870212343004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,7168,0.22037600146399605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,8192,0.2882933351728651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,7168,0.08626755740907456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,10240,0.2940639919704861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,6144,0.17035555839538574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,7168,0.24604622522989908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,5120,0.14344444539811876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,6144,0.0661262207561069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,6144,0.18030756049686006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,7168,0.20463821623060438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,4096,0.11163643995920818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,5120,0.05923110908932156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,5120,0.15865600109100342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,6144,0.18123911486731636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,3584,0.09720000293519761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,4096,0.04784799946679009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,4096,0.12059466044108073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,5120,0.15753155284457737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,3072,0.08805955780877008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,3584,0.04373777906099955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,3584,0.10739288727442424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,4096,0.12605599562327066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,2560,0.06747377581066556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,3072,0.03811644514401754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,3584,0.11275733841790093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,3072,0.09301777680714925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,2048,0.05714933077494303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,2560,0.03375555409325494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,1536,0.041265779071384005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,2560,0.08006844255659315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,3072,0.09825599855846828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,8192,0.23653599951002333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,1024,0.031832890378104314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,2048,0.06644533077875774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,2560,0.0856284432941013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,1536,0.024188444018363953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,768,0.02346311178472307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,1536,0.053393777873780995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,1024,0.020264888803164165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,1024,0.04000533289379544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,1536,0.05874311261706882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,512,0.017774222625626456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,768,0.018730666902330186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,768,0.03350399931271871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,1024,0.04649244414435493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,256,0.017831110292010836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,512,0.016926222377353244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,768,0.03946400019857619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,512,0.026354667213228013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,128,0.01550222271018558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,256,0.015492444237073263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,256,0.022985777921146814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,512,0.03256266646915012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,64,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,256,0.024745777249336243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,128,0.015824889143308003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,128,0.022294221652878657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,2048,0.02860355708334181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,128,0.01923644377125634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,64,0.022307554880777996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,32,0.021960000197092693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,2048,0.07158488697475858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,65536,0.5978275405036079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,32,0.014501333236694336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,65536,1.6158213085598414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,16384,0.35615110397338867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,51200,0.47177600860595703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,51200,1.713672849867079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,16384,0.15324444240993923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,16384,0.5607671207851833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,65536,2.3805698818630643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,12288,0.2749040126800537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,65536,1.7955421871609156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,51200,0.9866435792711047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,16384,0.40337422158983016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,12288,0.11217421955532497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,10240,0.24064087867736816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,12288,0.3196551005045573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,10240,0.08822755681143866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,8192,0.20160978370242646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,51200,1.3398515913221571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,12288,0.2936035527123345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,10240,0.29089776674906415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,8192,0.07120711273617215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,7168,0.18445155355665419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,10240,0.2450044420030382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,8192,0.2511262098948161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,7168,0.06460622284147474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,6144,0.1502515607410007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,5120,0.1248755587471856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,8192,0.20019200113084581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,7168,0.1961102220747206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,6144,0.05929955508973864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,4096,0.09416800075107151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,6144,0.18163555198245576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,7168,0.173927108446757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,5120,0.04842666784922282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,5120,0.12881955835554335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,3584,0.080713775422838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,6144,0.15370221932729086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,4096,0.04081955552101135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,4096,0.10477599832746719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,3072,0.07167910867267184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,3584,0.035563555028703474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,5120,0.13134755028618708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,3584,0.09175999959309895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,2560,0.05784800317552355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,3072,0.03257511059443156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,3072,0.08188800017038982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,2048,0.04848177896605598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,2560,0.02884711159600152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,2560,0.06959288650088839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,3072,0.0849368903372023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,1536,0.035972442891862653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,2048,0.0251528885629442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,2560,0.07414755556318495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,2048,0.058449778291914195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,1024,0.02533955540921953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,1536,0.021507556239763897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,2048,0.06343022320005628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,1536,0.046000887950261436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,768,0.019807111885812547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,1024,0.018246221873495314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,1024,0.03468622101677789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,1536,0.0522035559018453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,512,0.014487110906177096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,768,0.01679377754529317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,768,0.02886399957868788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,1024,0.040263109736972384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,256,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,512,0.0151573336786694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,768,0.03481511274973551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,128,0.010314666562610203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,4096,0.10729422171910603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,256,0.020287111401557922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,512,0.027912888262006972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,3584,0.09756977690590753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,64,0.008628444539176093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,32,0.008817777865462834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,256,0.022016000416543748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,128,0.014145778285132514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,128,0.019275556008021038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,64,0.019301333361201815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,128,0.01755911111831665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,32,0.01921599937809838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,65536,0.4939635594685872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,512,0.023174222972657945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,256,0.014112000664075216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,65536,1.3448275460137262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,51200,1.1115741729736328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,51200,0.3926240073310004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,51200,1.3478017383151586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,65536,1.7656977971394856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,16384,0.3578506575690375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,65536,1.5105875862969294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,16384,0.12110666433970134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,12288,0.24742844369676378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,16384,0.37607465849982363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,12288,0.09074399868647258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,10240,0.19546843899620903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,12288,0.24822754330105254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,51200,1.1397760179307725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,8192,0.17157422171698677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,10240,0.21151555909050834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,10240,0.09222666422526042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,12288,0.27042754491170246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,16384,0.3274426725175646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,7168,0.1452231142255995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,8192,0.06197600232230293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,8192,0.17736621697743735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,10240,0.21973600652482775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,6144,0.12807822227478027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,7168,0.06100533405939738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,7168,0.15017244550916883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,5120,0.10034844610426162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,6144,0.05126133229997423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,7168,0.15702933735317653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,6144,0.13162133428785536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,5120,0.042371554507149585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,4096,0.07488711012734307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,5120,0.1076328886879815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,6144,0.13760088549719915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,3584,0.05988444222344292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,4096,0.03397422366672092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,4096,0.08651110861036514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,3072,0.061305774582756885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,3584,0.03077955709563361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,3584,0.07670755518807305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,4096,0.09752444426218669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,2560,0.04692622356944614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,3072,0.02816177739037408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,3584,0.08733866612116496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,3072,0.06770844591988458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,2048,0.03839200072818332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,2560,0.025154666768179998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,2560,0.05829777982499865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,3072,0.07621066437827216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,8192,0.18266222212049696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,1536,0.030765334765116375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,2048,0.04837244417932299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,2048,0.02149066660139296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,2560,0.06613600254058838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,1024,0.02033333314789666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,1536,0.018600889378123812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,768,0.01721244388156467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,1536,0.0389120015833113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,1024,0.01583377851380242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,2048,0.057088891665140785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,512,0.01310400002532535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,1536,0.04584799872504341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,768,0.014825777875052558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,5120,0.11820799774593777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,768,0.024550222688251074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,256,0.009493333597977957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,1024,0.03639555639690823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,512,0.013591110706329346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,512,0.019959110352728102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,128,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,768,0.030455999904208716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,256,0.013067555924256643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,64,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,256,0.017159110969967313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,32,0.007551111280918121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,512,0.025243555506070454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,128,0.012990222209029727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,128,0.01646933290693495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,256,0.021078222327762183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,64,0.016526222229003906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,128,0.016861332787407767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,32,0.016177778442700703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,1024,0.029164443413416546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,65536,0.43585599793328184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,16384,0.2989448971218533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,65536,1.1162257724338107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,51200,0.8948710759480795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,51200,0.34810577498541934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,16384,0.12198400497436523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,12288,0.2091724475224813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,51200,1.1856702168782551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,65536,1.566688855489095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,16384,0.3876195483737522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,51200,1.0219706429375541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,12288,0.08090488778220283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,65536,1.3568124771118164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,12288,0.2395857705010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,8192,0.13557510905795628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,10240,0.07061155637105306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,10240,0.19144533740149605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,12288,0.2477733294169108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,8192,0.05719377597173055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,7168,0.1221039957470364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,8192,0.15130400657653809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,10240,0.21032710870107016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,6144,0.10492622190051609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,7168,0.05004622207747566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,7168,0.13367556201087102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,8192,0.17549599541558158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,5120,0.08327555656433105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,6144,0.04344800114631653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,7168,0.14310488435957167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,6144,0.11590311262342666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,4096,0.06549511353174846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,5120,0.03749422232309977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,10240,0.17843200100792778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,5120,0.09818755255805121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,3584,0.05170399943987528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,5120,0.10684266355302598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,3584,0.06972355312771268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,4096,0.0884408884578281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,3072,0.05086311035686069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,3584,0.02875111169285244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,3584,0.07923199733098348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,3072,0.06150311231613159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,3072,0.025864000121752422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,2560,0.04369866516855028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,3072,0.07060711251364814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,16384,0.33107466167873806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,2560,0.053328000836902194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,2048,0.03494400117132399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,2560,0.023279110590616863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,4096,0.030792888667848375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,2560,0.061407109101613365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,2048,0.020271999968422782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,2048,0.0443466669983334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,1024,0.019534221953815885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,2048,0.05112355616357592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,1536,0.017768000562985737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,1536,0.03555466731389364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,768,0.015170666906568738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,1024,0.015198222464985318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,1536,0.042797333664364286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,1024,0.026568889617919922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,512,0.011510222322411008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,768,0.02242044442229801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,768,0.014309333430396186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,1024,0.03339022066858079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,256,0.009874666730562845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,512,0.013086222112178802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,512,0.018358222312397428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,6144,0.12427022722032334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,128,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,256,0.012228444218635559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,4096,0.07826399803161621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,1536,0.0277128881878323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,64,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,256,0.019007111589113872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,128,0.015115555789735583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,32,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,64,0.01515111161602868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,32,0.01516533394654592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,768,0.028173334068722192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,65536,0.8612871170043945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,512,0.023171555664804246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,128,0.01240533341964086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,65536,0.39827556080288357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,128,0.01625066664483812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,65536,1.2115661833021376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,51200,0.7445671293470594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,16384,0.22217244572109648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,65536,1.1788542005750868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,51200,0.31495467821757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,16384,0.09410933653513591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,16384,0.27041421996222603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,12288,0.19849333498213026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,51200,0.9989173677232531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,256,0.015751111838552687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,10240,0.1666613287395901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,12288,0.20747466882069907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,16384,0.28170132637023926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,51200,0.8630737728542752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,10240,0.06834133466084798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,10240,0.1870488855573866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,12288,0.20919199784596762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,7168,0.11679555310143365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,8192,0.051516443490982056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,8192,0.13406133651733398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,10240,0.17615644137064615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,6144,0.09256444374720256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,7168,0.046480000019073486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,7168,0.11781333552466498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,8192,0.15160622861650255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,5120,0.07583377758661906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,6144,0.10131466388702393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,7168,0.13093600008222792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,4096,0.06286578045950995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,5120,0.0328435566690233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,6144,0.11259288258022732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,12288,0.07275022400750054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,3584,0.05268355541759067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,4096,0.027999109692043726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,5120,0.094851553440094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,4096,0.07017155488332112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,8192,0.13809688886006674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,3072,0.049379554059770375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,3584,0.0255084451701906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,4096,0.07851111226611666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,3584,0.06161599689059787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,2560,0.04105866617626614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,3072,0.022801778382725183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,3072,0.054487999942567616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,2560,0.020812445216708712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,6144,0.03826311230659485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,3072,0.06400799751281738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,2048,0.03295022249221802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,2560,0.054696887731552124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,5120,0.0871208906173706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,1536,0.02596355477968852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,2048,0.039637333816952176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,2048,0.047136889563666455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,1536,0.015983111328548856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,1024,0.017608889275126986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,1536,0.03171822097566392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,1024,0.014096000128322177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,1536,0.03847466574774848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,768,0.012902222573757172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,1024,0.023774221539497375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,512,0.010635555618339114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,3584,0.0712053312195672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,768,0.013139555851618448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,1024,0.029489778810077246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,768,0.020078221956888836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,2560,0.04683822062280443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,512,0.012326222327020435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,512,0.016540444559521146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,768,0.025496888491842482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,128,0.00665155549844106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,256,0.011545778148704104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,256,0.014184888866212634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,512,0.021278222401936848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,2048,0.018166222506099276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,64,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,32,0.006002666635645761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,128,0.011368888947698804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,128,0.01352355546421475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,256,0.018050667312410142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,64,0.013412444127930535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,32,0.013743999931547375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,65536,0.3231262101067437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,51200,0.48857869042290586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,65536,0.7517449061075846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,256,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,16384,0.2068106730779012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,51200,0.2557057804531521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,51200,0.7938559850056967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,65536,1.0158906512790256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,16384,0.22739643520779082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,12288,0.14417600631713867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,16384,0.08125778039296468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,65536,0.9571279949612088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,128,0.015045333239767285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,10240,0.12342045042249893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,12288,0.06654755274454753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,12288,0.18920089138878715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,16384,0.3054951032002767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,8192,0.10556177960501777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,10240,0.05453422334459093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,12288,0.1824773285124037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,10240,0.1569395595126682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,7168,0.08904444509082371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,8192,0.04367022381888496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,8192,0.11865688694847955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,6144,0.07671199904547797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,7168,0.03851288888189528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,7168,0.10306400060653687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,8192,0.12722310754987928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,6144,0.033518221643235944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,5120,0.05980088975694445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,7168,0.11352266205681695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,6144,0.08922400077184041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,4096,0.046966221597459584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,5120,0.02916533417171902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,6144,0.09841599729326035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,5120,0.07646755377451579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,3584,0.04098755452367995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,4096,0.024707555770874023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,4096,0.06061866548326281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,5120,0.08419289191563924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,3072,0.03762222329775492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,3584,0.022633777724372015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,3584,0.054620444774627686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,4096,0.06849866443210177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,2560,0.030976000759336684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,51200,0.7472115622626411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,3072,0.020584000481499564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,3584,0.062302225165896945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,2048,0.023711999257405598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,2560,0.018586667047606576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,10240,0.15496799680921766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,2560,0.04165777895185683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,1536,0.01891999940077464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,3072,0.05612622367011177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,2048,0.01662577854262458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,2048,0.03539199961556329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,1024,0.013754666679435305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,1536,0.014868444866604276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,2048,0.04222844375504387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,1536,0.02808977829085456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,768,0.011053333679835001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,1024,0.013047110703256396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,1024,0.021433777279324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,768,0.01808088852299584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,768,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,1024,0.026916444301605225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,768,0.023171555664804246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,512,0.014787554740905762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,3072,0.048235555489857994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,256,0.006450666735569636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,512,0.011122666299343109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,512,0.019588443968031142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,128,0.00620177760720253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,256,0.010679999987284342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,256,0.01276622215906779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,256,0.016931555337376065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,128,0.010776888993051318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,128,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,64,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,32,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,1536,0.0344888899061415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,128,0.01388711151149538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,64,0.012087110843923358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,32,0.012366222010718452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,512,0.008626666333940294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,65536,0.6674204402499728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,65536,0.2975280019972059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,51200,0.4987226592169867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,16384,0.1702337794833713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,51200,0.23610488573710123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,65536,0.8278586599561902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,51200,0.6459404627482096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,12288,0.13099911477830675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,16384,0.07805777920616998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,16384,0.2181831068462796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,65536,0.9262818230523003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,12288,0.06061866548326281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,10240,0.11286221610175239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,12288,0.14686755339304605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,16384,0.24194044537014434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,10240,0.05269066823853386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,8192,0.08629866441090901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,10240,0.1289528873231676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,12288,0.18200355105929902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,7168,0.0737288859155443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,8192,0.0395413339138031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,8192,0.10021866692437066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,10240,0.15343556139204237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,2560,0.048840890328089394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,7168,0.035595556100209556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,7168,0.08820266856087579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,8192,0.12544443872239855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,5120,0.05112089051140679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,6144,0.03110399842262268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,7168,0.10943377680248684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,6144,0.07717688878377278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,4096,0.042103111743927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,5120,0.02734222180313534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,5120,0.06457066535949707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,6144,0.09632800022761027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,3584,0.03657244311438666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,4096,0.05290844374232822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,4096,0.0233262214395735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,5120,0.08063022295633952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,3072,0.03257333238919576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,3584,0.021393777595625982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,4096,0.06849333312776354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,2560,0.028340445624457464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,3072,0.04201511210865445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,3584,0.06055021948284573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,3072,0.05440622236993578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,6144,0.06416177749633789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,2560,0.017893332574102614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,2560,0.036375999450683594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,2048,0.023870221442646448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,1536,0.019155555301242404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,2048,0.01588355501492818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,2560,0.0465191106001536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,2048,0.030918220678965252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,1024,0.01386755539311303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,1536,0.02472266720400916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,1536,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,2048,0.039228445953793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,768,0.011688888900809817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,1024,0.01260355528857973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,1024,0.018743111027611625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,1536,0.03185688787036472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,512,0.008977777428097194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,51200,0.7298613124423556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,768,0.011975111232863532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,1024,0.025751110580232408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,256,0.006396444307433234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,3072,0.019495111372735765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,512,0.011015111373530494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,512,0.013131555583741931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,768,0.021707556313938562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,128,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,256,0.010290666586822933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,256,0.011379555695586734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,512,0.018862222631772358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,64,0.006078222145636876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,256,0.01570666664176517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,32,0.006487111250559489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,64,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,128,0.01295555548535453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,32,0.01074222226937612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,768,0.015857777661747403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,65536,0.4951493475172255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,65536,0.24693332778082955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,65536,0.6654568778143989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,51200,0.38368799951341415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,128,0.01068800025516086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,16384,0.12521778212653265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,51200,0.2138000064425998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,65536,0.7693475617302789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,51200,0.5239289071824815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,16384,0.06502844227684869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,16384,0.16186843978034124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,12288,0.09727911154429118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,12288,0.05245777633455065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,51200,0.6064142121209039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,16384,0.21036622259351942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,12288,0.12354755401611328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,12288,0.15410399436950684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,10240,0.10453244712617661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,128,0.010732444624106089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,8192,0.06596977843178643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,7168,0.05464355481995476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,8192,0.033598221010631986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,10240,0.13013511233859593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,8192,0.08381688594818115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,6144,0.045347554816140064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,7168,0.030415111117892798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,7168,0.07375288671917386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,8192,0.10832089185714722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,5120,0.03776977790726556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,6144,0.026872888207435608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,6144,0.06433688931994967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,7168,0.0937102238337199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,4096,0.0292871097723643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,5120,0.05456533365779453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,10240,0.08400799830754598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,6144,0.08167466852400038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,3584,0.02498311135503981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,4096,0.01994400057527754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,10240,0.04452711012628344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,4096,0.04509600003560384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,5120,0.07047466437021892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,3072,0.021199110481474135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,3584,0.01847555571132236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,3584,0.039946665366490684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,4096,0.05916355715857612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,2560,0.020083554916911654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,3072,0.03549422158135308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,3584,0.05271111263169182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,2048,0.016237333416938782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,3072,0.046559111939536206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,2560,0.015536889433860779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,3584,0.04774311184883118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,1536,0.01332800007528729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,2560,0.040046221680111356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,2048,0.01443822185198466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,2048,0.02622844444380866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,1024,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,1536,0.012680889003806643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,5120,0.023518222901556227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,2048,0.034749332401487566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,768,0.007833778030342525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,1024,0.011346666349305047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,1024,0.016198222835858662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,1536,0.028527110815048218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,512,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,768,0.010786666638321347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,768,0.013759999639458127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,1024,0.022867555419603985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,256,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,512,0.009995555712117089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,512,0.011735999749766456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,3072,0.017121778594122995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,768,0.020115555988417733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,128,0.004860444615284602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,256,0.009986666341622671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,2560,0.030584888325797185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,512,0.01686755485004849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,64,0.004256000121434529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,32,0.004219555606444676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,128,0.009322666459613377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,128,0.009649777577983009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,256,0.015115555789735583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,128,0.01144266708029641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,64,0.009353777600659264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,32,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,65536,0.2982737753126356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,65536,0.19559200604756674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,51200,0.2542400095197889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,65536,0.5151848793029785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,51200,0.15532178348965114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,51200,0.40415112177530926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,16384,0.09201066361533271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,256,0.009704888694816166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,65536,0.6221511099073622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,16384,0.053351110882229276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,12288,0.06400089131461249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,51200,0.49138932757907444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,16384,0.13201510906219482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,10240,0.05265688896179199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,12288,0.04209511147605049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,16384,0.16614222526550293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,12288,0.10156266556845771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,8192,0.0417866673734453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,10240,0.08409777614805435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,10240,0.03696977761056688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,12288,0.12838133176167807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,7168,0.036074668169021606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,8192,0.02731200059254964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,10240,0.10859644412994385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,8192,0.0680604444609748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,6144,0.029288000530666772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,7168,0.024792000651359558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,7168,0.05988978015051948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,8192,0.08929066525565253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,5120,0.026230222649044458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,6144,0.021873777111371357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,6144,0.051868445343441434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,7168,0.07838844590716891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,4096,0.020766221814685397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,5120,0.019274666905403137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,3584,0.01906044450071123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,5120,0.04422577884462145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,6144,0.06884088781144884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,4096,0.01682844426896837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,3072,0.016547555724779766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,4096,0.03674755493799845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,5120,0.05918666389253405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,3584,0.01569955547650655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,2560,0.014759999182489185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,4096,0.049842665592829384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,3072,0.014512000812424554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,1536,0.02149244480662876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,2048,0.012079111403889127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,3584,0.04451555675930447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,2560,0.013441777891582914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,3072,0.03931466738382975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,1536,0.010411555568377176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,2048,0.012096000214417776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,2048,0.021664000219768945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,2560,0.034007999632093645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,1024,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,2048,0.029161777761247423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,1536,0.011108444796668159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,1536,0.017679999272028606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,768,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,1024,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,1024,0.013591110706329346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,768,0.011709333293967776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,512,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,768,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,3584,0.0333315564526452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,768,0.01791200041770935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,3072,0.029457777738571167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,256,0.004570666700601578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,512,0.00997866690158844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,512,0.009316444396972656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,2560,0.025428444147109985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,512,0.015139554937680563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,256,0.008646222452322641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,256,0.008608889248636033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,64,0.003552888830502828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,128,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,256,0.013172444370057849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,32,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,64,0.007875555919276344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,128,0.010760889285140567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,1536,0.024404444628291663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,1024,0.02000177734427982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,65536,0.1914808882607354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,51200,0.20362933476765951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,65536,0.43105422125922305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,128,0.003899555653333664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,16384,0.06603288650512695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,51200,0.15285066763559976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,128,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,51200,0.34370843569437665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,65536,0.6202898025512695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,12288,0.04831288920508491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,16384,0.11254399352603489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,65536,0.24211377567715117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,12288,0.0854862199889289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,12288,0.04168000155025058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,10240,0.038408888710869685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,51200,0.48796357048882383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,16384,0.16483822133806017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,8192,0.030292444758945044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,12288,0.1272746721903483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,10240,0.03633244501219855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,10240,0.07195378012127347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,8192,0.058006220393710665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,7168,0.026513778501086768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,8192,0.026869333452648584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,10240,0.10817511214150323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,7168,0.024168888727823894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,7168,0.05139999919467502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,7168,0.07756977611117892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,6144,0.04515022370550367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,5120,0.019091554813914828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,16384,0.052296890152825244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,6144,0.02160266704029507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,32,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,6144,0.06810666455162896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,5120,0.01926488843229082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,5120,0.03833511140611436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,5120,0.05760266383488973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,4096,0.031832890378104314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,3584,0.01387644476360745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,4096,0.048250665267308555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,8192,0.08788266446855332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,3584,0.028480887413024902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,3072,0.012416000167528788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,3584,0.015513777732849121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,3584,0.04305600126584371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,2560,0.011223999990357293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,3072,0.014463111758232117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,3072,0.025285333395004272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,3072,0.03830755419201321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,2560,0.013280889226330651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,2048,0.010015111002657149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,2560,0.02199733257293701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,4096,0.015504888362354703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,1536,0.00815466625822915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,2048,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,2560,0.03322488731808133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,1024,0.006322666588756774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,2048,0.01902844508488973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,4096,0.01686755485004849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,1536,0.011028444601429833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,768,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,2048,0.028749333487616643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,1024,0.010261333651012844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,1024,0.012250666817029318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,512,0.004549333204825719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,6144,0.02478755513827006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,768,0.009737778041097853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,768,0.010630221830474006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,1024,0.01924533314175076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,512,0.009093333449628618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,768,0.017522666189405654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,512,0.009170666337013245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,128,0.003478222423129611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,256,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,512,0.015065777632925244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,128,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,256,0.013076444466908773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,64,0.0031928889867332247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,32,0.00318755561278926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,128,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,64,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,1536,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,32,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,65536,0.18968978193071154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,1536,0.02407199972205692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,51200,0.15339822239345974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,256,0.0038417776425679526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,65536,0.3850444422827826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,51200,0.11660444736480713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,16384,0.049117333359188504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,256,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,51200,0.30187998877631295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,12288,0.03797066542837355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,65536,0.47897953457302517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,128,0.00871377769443724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,16384,0.041379554404152766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,16384,0.10038755337397258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,10240,0.031055112679799397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,51200,0.3842657672034369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,12288,0.03263288736343384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,12288,0.07654577493667603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,8192,0.023409777217441138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,16384,0.13072800636291504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,10240,0.028233776489893597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,7168,0.020615110794703167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,10240,0.06430577569537692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,12288,0.10080089171727498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,8192,0.020247111717859905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,6144,0.018340445227093168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,8192,0.051464888784620494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,10240,0.08498933580186631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,7168,0.018244443668259513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,5120,0.015885333220163982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,8192,0.06953866614235772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,6144,0.016506666938463848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,6144,0.04024799995952182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,4096,0.013291555146376291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,7168,0.06077510780758328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,5120,0.014800000521871777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,5120,0.034183998902638756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,6144,0.05407911207940844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,3584,0.011955555942323474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,4096,0.028370665179358587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,4096,0.013110222087966071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,5120,0.045088890526029796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,3072,0.010572444233629437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,4096,0.03822044531504313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,3584,0.025589333640204534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,3072,0.011769777370823754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,3072,0.022852443986468848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,2560,0.009273777405420939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,3584,0.03436000148455302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,65536,0.1456000010172526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,3072,0.030643555853101943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,2560,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,7168,0.04623200164900886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,2048,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,1536,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,2048,0.01722666621208191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,2048,0.023765333824687537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,1536,0.014195554786258273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,1024,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,1536,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,3584,0.012451555165979596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,1536,0.019946666227446664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,768,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,1024,0.011149333582984077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,768,0.008711111214425828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,1024,0.016859556237856548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,768,0.009653333160612319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,512,0.004202666795916027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,2560,0.020131554868486192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,768,0.014834667245546976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,256,0.0035217776894569397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,512,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,2560,0.027461333407296076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,128,0.003220444545149803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,512,0.012759110993809171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,256,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,128,0.008021333151393467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,256,0.011712888876597086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,128,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,64,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,32,0.0028524444335036804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,65536,0.11195111274719238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,128,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,1024,0.009200000100665623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,64,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,32,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,51200,0.09053511089748806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,65536,0.14495910538567439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,2048,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,65536,0.3111306561364068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,512,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,51200,0.24400356080796984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,51200,0.11572710673014323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,12288,0.025918222135967676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,16384,0.04056355688307021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,65536,0.4569519890679254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,16384,0.08185866806242201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,10240,0.02035911050107744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,51200,0.36372711923387313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,12288,0.06194133228725857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,12288,0.03224533465173509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,16384,0.12625155184004042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,8192,0.0163093325164583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,10240,0.027971555789311726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,10240,0.05201511250601875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,12288,0.09746488597657944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,8192,0.042352000872294106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,8192,0.020059555768966675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,10240,0.08272177643246122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,8192,0.066012442111969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,7168,0.017878222796652053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,7168,0.03791911072201199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,16384,0.033542222446865506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,6144,0.01648355523745219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,5120,0.011398221883508893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,7168,0.05962489048639933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,6144,0.033391998873816595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,4096,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,256,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,5120,0.014632888966136508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,6144,0.05204711026615567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,3584,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,5120,0.028299556838141546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,4096,0.01312355531586541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,4096,0.023410666320059035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,3072,0.00758844448460473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,3584,0.01239377756913503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,3584,0.02128444446457757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,4096,0.03667022122277154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,7168,0.014799111419253878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,2560,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,3584,0.03331999977429708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,6144,0.013057777451144325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,2560,0.016784888174798753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,2560,0.011075555450386472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,2048,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,3072,0.029951999584833782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,2560,0.02609955602222019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,1536,0.005192888693677055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,2048,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,2048,0.014508444401952954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,2048,0.022999111149046157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,1536,0.009688888986905416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,1024,0.004224888980388641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,1536,0.012072000238630505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,1024,0.00908622228437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,5120,0.04427822099791633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,768,0.0038693332009845306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,1024,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,768,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,1024,0.015864888827006023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,512,0.0031866667171319327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,768,0.008669333325492011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,3072,0.018982221682866413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,256,0.0031786666562159858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,768,0.014489778214030795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,512,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,512,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,128,0.0028391112056043413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,256,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,512,0.012746666868527731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,128,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,256,0.011037333144081963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,64,0.0026542221506436667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,32,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,128,0.008401778009202745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,1536,0.019367999500698514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,64,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,65536,0.08307733138402303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,32,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,51200,0.06756799750857882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,3072,0.011704000333944956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,65536,0.1438328954908583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,65536,0.26785511440700954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,16384,0.026142223013771906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,51200,0.11499110857645671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,51200,0.2106177806854248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,256,0.007985778152942657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,12288,0.017860444055663217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,65536,0.45629511939154727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,16384,0.04044977823893229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,16384,0.0704924464225769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,10240,0.015423999892340766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,12288,0.05384000142415365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,51200,0.3593635559082031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,8192,0.01293777757220798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,16384,0.12348888980017768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,10240,0.028692444165547688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,10240,0.045440001620186694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,12288,0.09640444649590386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,7168,0.011739555332395764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,8192,0.02056799994574653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,10240,0.08090222544140287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,8192,0.03673155440224542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,6144,0.010278222461541494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,7168,0.0192284451590644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,7168,0.03282489048110114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,8192,0.0652995573149787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,5120,0.0087013335691558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,6144,0.028872887293497723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,6144,0.01721244388156467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,7168,0.05880622069040934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,4096,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,6144,0.05114755696720547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,5120,0.02488533324665493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,3584,0.006565333240562015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,4096,0.020748444729381137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,4096,0.013125333521101209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,5120,0.04350577791531881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,3072,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,12288,0.032772445016437106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,4096,0.03674755493799845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,3584,0.012448000411192576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,3584,0.018883556127548218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,2560,0.005481777919663324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,3072,0.012062222593360476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,3072,0.016896888613700867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,2048,0.00453333349691497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,3584,0.032616890139049955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,2560,0.01513155632548862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,2560,0.011399110986126794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,3072,0.029373334513770208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,1536,0.0042071110672420925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,2048,0.01035911093155543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,2048,0.013185777597957186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,1024,0.0033440000067154565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,1536,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,1536,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,2048,0.022674666510687932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,768,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,5120,0.014846222268210517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,1536,0.01896711190541585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,1024,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,512,0.0027919999427265594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,1024,0.008995555341243744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,128,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,768,0.008824889030721452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,768,0.008231999973456064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,256,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,1024,0.015503111812803479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,512,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,512,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,128,0.0021991110924217436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,768,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,256,0.00812355594502555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,512,0.012522666818565793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,256,0.005712000032265981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,64,0.0021484444538752237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,32,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,128,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,128,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,256,0.01032266683048672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,65536,0.07484710878796048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,51200,0.06112533145480686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,64,0.005281777845488654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,128,0.008299555215570662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,16384,0.02428177826934391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,32,0.005427555491526921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,12288,0.013920888304710388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,10240,0.012638222012254926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,8192,0.013567999833159976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,65536,0.26780443721347386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,16384,0.07020622491836548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,7168,0.010834666589895884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,51200,0.2101973295211792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,10240,0.0451493329472012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,5120,0.008295999632941352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,8192,0.03672177924050225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,4096,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,7168,0.03254844413863288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,6144,0.028519110547171697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,3584,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,3072,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,5120,0.02465155555142297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,2560,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,4096,0.02064622276359134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,2048,0.004550222307443619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,3584,0.01880088945229848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,1536,0.003855111284388436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,3072,0.016944888565275405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,1024,0.0032044444233179092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,2560,0.025811556312772963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,2048,0.013182222015327878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,768,0.0028728888266616394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,512,0.0026782221264309357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,1536,0.011056889262464313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,256,0.0022906667242447534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,1024,0.009011555876996782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,128,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,768,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,64,0.002151111140847206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,12288,0.05352177884843615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,512,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,32,0.0021893332401911416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,6144,0.009339555270142024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,65536,0.07486133442984687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,128,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,16384,0.020576889316240948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,12288,0.01219466659757826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,64,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,10240,0.010439111126793755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,65536,0.2689102225833469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,16384,0.07021244366963704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,8192,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,12288,0.05367822117275662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,7168,0.009005332986513773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,6144,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,10240,0.04522666666242811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,5120,0.010681777364677854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,2560,0.014946666028764514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,4096,0.008849778109126622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,8192,0.03668533431159125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,7168,0.03262933426433139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,3584,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,3072,0.007040888898902469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,6144,0.028554668029149372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,2560,0.0063484443558586975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,5120,0.024720888998773363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,2048,0.004535111288229625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,4096,0.02075822154680888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,1536,0.003846222327815162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,3584,0.018940443793932598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,1024,0.0032133333798911837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,2560,0.015174221661355762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,51200,0.05943466557396782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,768,0.0028782222006056043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,512,0.0025315555847353404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,2048,0.013252444565296173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,1536,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,256,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,32,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,128,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,1024,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,64,0.0018648888087934917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,32,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,768,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,512,0.006990222467316522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,51200,0.2099813355339898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,256,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,128,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,64,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,32,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,3072,0.017147555947303772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,16384,1.471343994140625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,12288,1.0406328837076824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,12288,3.202647950914171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,16384,4.241817050509982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,16384,4.800159878200955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,12288,3.340215894911024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,10240,0.8709120220608182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,51200,5.471304999457465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,8192,0.7171404626634387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,10240,2.7362986670600047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,10240,3.4947280883789062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,8192,2.1135457356770835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,8192,2.3372472127278647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,7168,0.6405990918477377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,6144,0.5352320141262478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,7168,1.966973410712348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,6144,1.793100357055664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,7168,2.252424875895182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,6144,1.5548844867282445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,5120,1.4957964155409071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,5120,0.47407113181220156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,4096,0.3910515573289659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,5120,1.3797208997938368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,4096,1.1345546510484483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,3584,0.3451591067843967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,3584,1.0124657948811848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,4096,1.1317742665608723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,3072,0.7952426804436578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,3584,0.9755635791354709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,2560,0.6974088880750867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,2560,0.26240889231363934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,3072,0.7942070960998535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,2048,0.5637049145168728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,2560,0.6751324335734049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,2048,0.21924355294969347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,1536,0.38270754284328884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,1536,0.17784621980455187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,1024,0.2917822202046712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,3072,0.3244746790991889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,1024,0.1409511168797811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,1536,0.5017929077148438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,768,0.2294222248925103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,1024,0.33869510226779515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,512,0.1884453296661377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,768,0.12271289030710857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,256,0.13331554995642766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,768,0.25553154945373535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,512,0.11214133103688557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,128,0.0899511112107171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,512,0.2562088966369629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,256,0.10479199886322021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,51200,13.015628390842013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,256,0.19834666781955293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,64,0.07731822464201185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,32,0.0994559990035163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,128,0.095195558336046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,128,0.18705066045125326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,64,0.19462311267852783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,32,0.1928257809744941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,2048,0.609940422905816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,16384,3.273076375325521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,51200,4.1750844319661455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,65536,5.556323581271702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,16384,1.107450697157118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,51200,20.577729966905384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,16384,3.6490754021538625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,12288,0.8345973226759168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,10240,2.13570319281684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,12288,2.8862516615125866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,51200,10.093133714463976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,12288,2.8829839494493275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,10240,0.6715920236375598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,8192,0.5480888684590658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,10240,2.273193783230252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,8192,1.6494586732652452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,7168,0.49249956342909074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,8192,1.7004835340711806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,7168,1.6705111397637262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,7168,1.5374675326877172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,65536,12.93022240532769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,6144,0.4272933271196153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,6144,1.4914675818549261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,6144,1.2965173721313477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,5120,1.1791138119167752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,5120,0.36078845130072695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,4096,0.9088293711344401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,4096,0.30204089482625324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,5120,1.030417760213216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,3584,0.27076178126864964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,4096,0.8619750870598687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,3584,0.7776231235928006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,3584,0.7391111056009928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,3072,0.7154391076829699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,3072,0.2354551156361898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,3072,0.6234409014383951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,2560,0.5316230985853407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,2560,0.24229065577189127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,2048,0.45846043692694766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,2048,0.18774844540490043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,2560,0.6129057672288682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,2048,0.4271404478285048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,1536,0.37833333015441895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,1536,0.14442755116356745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,1024,0.25572088029649526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,768,0.18362132708231607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,1024,0.1286497778362698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,1536,0.42040443420410156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,512,0.1438266701168484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,768,0.10266933176252578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,1024,0.2557057804531521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,256,0.0986986690097385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,512,0.09616177611880833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,512,0.17674044768015543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,128,0.0823048883014255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,256,0.07976000176535712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,256,0.1554888884226481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,64,0.062112000253465444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,128,0.08378044764200847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,51200,16.54384019639757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,32,0.06497689088185628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,128,0.14737955729166666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,64,0.1498862240049574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,32,0.15415199597676596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,768,0.22820533646477592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,16384,1.2631813685099285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,51200,1.285080909729004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,65536,1.7914098103841145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,16384,0.4517546759711371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,12288,0.9275173611111112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,16384,1.22926754421658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,51200,3.3850951724582248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,12288,0.32716088824801975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,12288,0.8825519879659017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,10240,0.22761776712205675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,10240,0.8199146588643392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,8192,0.5175955560472276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,65536,4.505710177951388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,10240,0.7322373390197754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,8192,0.22691734631856283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,7168,0.5042844348483616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,8192,0.5789155430263943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,7168,0.16649688614739314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,6144,0.45560889773898655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,6144,0.14510133531358507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,5120,0.36228354771931964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,51200,5.210150400797526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,7168,0.506424003177219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,5120,0.12516178025139704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,6144,0.4613688786824544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,4096,0.31760888629489475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,65536,23.01529608832465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,4096,0.10453599691390991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,3584,0.25592266188727486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,3072,0.2233048809899224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,3584,0.10542666912078857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,4096,0.30197866757710773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,3072,0.08225511180029975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,3584,0.24237778451707628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,3072,0.20879467328389487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,2048,0.1447395616107517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,2560,0.07375555568271212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,1536,0.10781333181593154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,2048,0.06207999918195936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,2560,0.1894151104821099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,2048,0.1457093291812473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,1024,0.07659910784827338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,1536,0.050705777274237744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,768,0.06336266464657254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,1536,0.11628354920281304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,512,0.046373334195878774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,1024,0.08912533521652222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,768,0.038423998488320246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,65536,7.3177744547526045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,256,0.03152889013290405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,768,0.07298044363657634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,512,0.035291555855009295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,128,0.0262746661901474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,5120,0.40498489803738064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,256,0.031593776411480375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,64,0.023344000180562336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,32,0.023103111320071753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,256,0.0675742228825887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,128,0.030917333232031927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,128,0.051586667696634926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,2560,0.18350756168365479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,64,0.0521982212861379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,32,0.05282488796446058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,1024,0.042223999897638954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,512,0.06064089139302572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,16384,0.6859511269463433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,51200,0.9856657452053494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,65536,1.2564516067504883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,16384,0.2687973446316189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,16384,0.898628446790907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,12288,0.7003253300984701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,12288,0.2133208910624186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,51200,2.5179191165500217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,12288,0.6653422249688042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,10240,0.18821333514319527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,10240,0.5498426755269369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,65536,3.1188151041666665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,8192,0.36454934544033474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,10240,0.533380455440945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,8192,0.17135022746192086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,7168,0.3752888838450114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,7168,0.1303644445207384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,8192,0.4310035440656874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,6144,0.40863466262817383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,7168,0.3848373360104031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,6144,0.11379644605848525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,5120,0.23331112331814238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,51200,3.9183875189887156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,4096,0.18612089422014025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,5120,0.09525866640938653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,6144,0.3544159995185004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,4096,0.08446400033103095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,3584,0.17205511199103463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,3584,0.0714248882399665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,3072,0.16000978151957193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,4096,0.22641423013475206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,3584,0.18482400311364067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,3072,0.06426755587259929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,2560,0.05688799752129448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,2048,0.1054133309258355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,3072,0.16559822029537624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,1536,0.07449866665734185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,2560,0.13778400421142578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,65536,5.115440792507596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,2048,0.049294223388036094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,1024,0.05854666895336575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,2048,0.11281066470675999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,768,0.045406222343444824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,1024,0.03444088829888238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,1536,0.08922488821877374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,512,0.03702577617433336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,768,0.03139644530084398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,1024,0.08741955624686347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,256,0.024675556355052527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,768,0.055323554409874805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,128,0.019118222925398085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,512,0.02865155537923177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,512,0.0473235547542572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,256,0.025972444150182936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,64,0.016141333513789706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,5120,0.2961600091722277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,256,0.046000000503328115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,128,0.025230222278171118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,128,0.03995555639266968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,2560,0.13965422577328152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,64,0.040531555811564125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,32,0.04111999935574002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,1536,0.04024799995952182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,32,0.01643822259373135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,16384,0.5729342036777073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,51200,0.8194177945454916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,65536,1.0834230846828885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,16384,0.23046665721469453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,12288,0.17732000350952148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,12288,0.7065893279181586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,51200,2.0328701866997614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,16384,0.7501875559488932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,10240,0.38567201296488446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,8192,0.31391999456617564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,12288,0.8278382089402941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,10240,0.1497511068979899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,8192,0.12357866764068604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,7168,0.2789822154574924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,10240,0.5490293502807617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,7168,0.31554489665561253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,7168,0.10998221900728013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,8192,0.3874026669396295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,6144,0.24121244748433432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,5120,0.1965022219551934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,6144,0.0961644450823466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,6144,0.26808889706929523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,4096,0.16415022479163274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,5120,0.23689511087205675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,4096,0.06878399848937988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,3584,0.1430799961090088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,4096,0.17848089006212023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,65536,2.657022264268663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,3584,0.061958220269944936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,3072,0.12814577420552573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,3584,0.15774310959710017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,2560,0.11121955182817246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,3072,0.05491022268931071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,65536,4.340352800157335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,2048,0.09218666950861613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,3072,0.13611200120713976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,2560,0.049024889866511025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,1536,0.07202044460508558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,2560,0.11798844072553848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,1024,0.04900799857245552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,1536,0.03536533315976461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,2048,0.048230220874150596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,1536,0.07632088661193848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,768,0.036079999473359846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,51200,3.1672106848822703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,1024,0.028765334023369685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,512,0.02955377764172024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,1024,0.06042844719356961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,256,0.020181333025296528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,768,0.027064000566800434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,512,0.024061333802011278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,768,0.04702755477693346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,128,0.017619555195172627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,512,0.03926577832963731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,64,0.014789332946141561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,256,0.03479377759827508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,128,0.020920889245139226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,32,0.014836443795098199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,128,0.032736000087526106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,64,0.03273333443535699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,32,0.032467554012934365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,2048,0.09595288832982381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,5120,0.08145866791407268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,256,0.02253866692384084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,65536,0.8700933456420898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,16384,0.4647253354390462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,51200,0.663499567243788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,12288,0.3816568851470947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,16384,0.24650666448805067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,16384,0.6019742223951552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,51200,1.6499421861436632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,12288,0.14572889275021023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,65536,2.2436230977376304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,10240,0.3829190995958116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,10240,0.135827554596795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,12288,0.5031093491448296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,10240,0.36359556516011554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,8192,0.284496890174018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,8192,0.11970310741000706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,7168,0.2844613393147786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,7168,0.09042044480641682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,6144,0.21389333407084146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,8192,0.31931376457214355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,7168,0.26173244582282174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,6144,0.10312444633907741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,5120,0.20895732773674858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,6144,0.23530843522813585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,51200,2.4520062340630426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,4096,0.14981510904100206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,3584,0.12339911195966934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,4096,0.056623107857174344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,5120,0.19838310612572563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,3072,0.1172986692852444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,3584,0.052616000175476074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,65536,3.338219536675347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,4096,0.17045956187778047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,2560,0.08076088958316378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,3072,0.046781334612104625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,2048,0.07243822018305461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,3584,0.13626489374372694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,2560,0.04481688804096646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,3072,0.12496266100141738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,1536,0.060306668281555176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,2048,0.035410665803485446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,1024,0.04140533341301812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,768,0.030574222405751545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,1536,0.02996888756752014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,2048,0.08521511157353719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,1024,0.024163555767801072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,1536,0.06292444467544556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,1024,0.048808889256583325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,768,0.022100443641344707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,256,0.01648355523745219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,512,0.019595555133289762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,128,0.014452444182501899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,256,0.01797777745458815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,5120,0.06848266389634874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,256,0.02767466670937008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,64,0.012084444363911947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,32,0.012376888758606382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,128,0.017858665850427415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,128,0.02682133350107405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,64,0.026733333865801494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,32,0.026425777210129633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,2560,0.09891821940739949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,512,0.02372088862790002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,768,0.041646222273508705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,512,0.0317884451813168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,65536,0.8605928950839572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,16384,0.504925356970893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,51200,0.6811582247416178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,51200,1.5167528788248699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,16384,0.5292497740851508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,16384,0.23589865366617838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,12288,0.3281875451405843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,65536,1.9530631171332464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,12288,0.13870222038692898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,10240,0.12643999523586696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,8192,0.25005332628885907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,12288,0.4842835532294379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,10240,0.3382355637020535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,8192,0.09835377666685317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,7168,0.21330044004652235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,8192,0.2569004429711236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,7168,0.08250044451819526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,51200,2.236349317762587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,6144,0.17758133676317003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,7168,0.22563555505540636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,6144,0.07250133487913343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,5120,0.15504444970024958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,65536,3.028601752387153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,6144,0.1950204504860772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,5120,0.06269244352976482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,3584,0.10517688592274983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,5120,0.16996355851491293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,4096,0.06324266725116305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,10240,0.2608017656538221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,3072,0.09122133255004883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,4096,0.14936533239152697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,3584,0.04758488800790575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,3584,0.1143511136372884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,3072,0.04169599877463447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,2048,0.060862223307291664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,2560,0.03644088904062907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,3072,0.10477599832746719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,1536,0.05046311020851135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,2048,0.030925333499908447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,2560,0.08851910961998834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,1024,0.03456444541613261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,1536,0.026064000195927087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,2048,0.07123021947013007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,768,0.027386666999922857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,1536,0.056440889835357666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,1024,0.02146311104297638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,512,0.01975822283162011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,1024,0.044222222434149853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,768,0.01943733294804891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,256,0.014730667074521383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,512,0.018008889423476327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,768,0.03560533457332187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,128,0.012198222180207571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,512,0.028532445430755615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,256,0.016532444291644625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,64,0.010604444477293225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,32,0.011219555305110084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,256,0.02475999957985348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,128,0.016280000408490498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,128,0.023771555887328252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,64,0.02400355537732442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,2560,0.08329866992102729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,32,0.02370044423474206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,16384,0.4441537857055664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,51200,0.5586613549126519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,65536,0.7241795327928332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,4096,0.12226666344536676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,16384,0.17172000143263075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,16384,0.4747173521253798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,12288,0.4076480070749919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,12288,0.33515021536085343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,12288,0.12307911449008518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,65536,1.865121841430664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,10240,0.2741813394758436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,8192,0.20346577962239584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,10240,0.10105955600738525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,8192,0.08130577537748548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,7168,0.17409866385989717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,10240,0.3545111020406087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,6144,0.15582666132185194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,8192,0.24148800637986925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,7168,0.07305422094133165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,7168,0.19585866398281523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,5120,0.14464088280995688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,6144,0.07196799914042155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,5120,0.05498399999406603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,6144,0.17356533474392363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,4096,0.11829333835177952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,3584,0.08087911208470662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,5120,0.14374489254421657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,4096,0.046118223004870944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,3072,0.0834231111738417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,51200,1.471845308939616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,4096,0.11972978379991318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,3584,0.0418542226155599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,2560,0.06403377983305189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,3584,0.10552533467610677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,3072,0.03695733348528544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,2048,0.05189777745140923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,2560,0.03295822276009454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,3072,0.08835999833212958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,1536,0.0389048887623681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,1024,0.026492445005310908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,2560,0.07653511232799955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,1536,0.02348711093266805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,51200,1.903707504272461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,768,0.02141422198878394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,1024,0.02055911057525211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,1536,0.05009066727426317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,512,0.01667022208372752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,1024,0.037951111793518066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,768,0.01830222209294637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,256,0.01255644443962309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,512,0.016619554824299283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,65536,2.6665965186225042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,512,0.02520622147454156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,128,0.009898666706350114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,256,0.022044445077578228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,128,0.015536889433860779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,128,0.02128533356719547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,32,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,64,0.02127733329931895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,2048,0.02776088813940684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,32,0.020982222424613103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,2048,0.06428711281882392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,768,0.03167288833194309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,256,0.01516177753607432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,64,0.008798222574922774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,65536,0.5781964196099175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,65536,1.308916409810384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,16384,0.325018670823839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,51200,1.125718222724067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,16384,0.1389804416232639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,51200,0.46599377526177305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,12288,0.2745511002010769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,12288,0.1269342237048679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,16384,0.4473315344916449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,12288,0.3065759870741102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,8192,0.20690311325920951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,10240,0.10547200176450942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,10240,0.25940177175733775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,8192,0.07318933142556085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,7168,0.18426667319403756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,8192,0.19389866458045113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,7168,0.0646373364660475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,65536,2.1267378065321183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,51200,1.577652401394314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,5120,0.11294133133358425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,7168,0.1765128903918796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,6144,0.05580177572038439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,4096,0.08046488629447089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,5120,0.049920890066358775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,6144,0.143577774365743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,3584,0.06910577747556898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,4096,0.03904089000489976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,3072,0.06569955746332805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,4096,0.0985564457045661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,3584,0.035120887888802424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,2560,0.05800266398323906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,3072,0.031870222753948636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,3584,0.08989333444171482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,2048,0.04454666707250807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,10240,0.22898401154412162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,3072,0.0755315555466546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,1536,0.035387555758158364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,2560,0.028502222564485338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,2560,0.0692053304778205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,1024,0.023982221881548565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,2048,0.024774221910370722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,768,0.018438221679793466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,2048,0.054604444238874644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,1536,0.020997333857748244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,512,0.013882666826248169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,1536,0.043488889932632446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,768,0.016171556380059984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,256,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,768,0.027884443600972492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,6144,0.128255115614997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,512,0.014897776974572076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,128,0.00941955546538035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,512,0.02200177808602651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,256,0.013779555757840475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,64,0.008294222255547842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,32,0.008634666601816813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,256,0.019294222195943195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,128,0.018573333819707234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,5120,0.12935377491845026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,64,0.01852799952030182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,32,0.01847111185391744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,1024,0.018166222506099276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,65536,0.4938977559407552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,1024,0.03349155518743727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,51200,0.7569706704881457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,65536,1.0200364854600694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,16384,0.3174897829691569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,12288,0.21330311563279894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,128,0.014083556003040738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,51200,0.3865537908342149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,16384,0.3159831100040012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,10240,0.18061955769856772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,12288,0.0992471112145318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,8192,0.1444666650560167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,51200,1.294865820142958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,12288,0.3272248903910319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,65536,1.6087599860297308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,10240,0.2301830980512831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,7168,0.135261336962382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,8192,0.06127822399139404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,6144,0.11947911315494114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,8192,0.16544444031185573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,7168,0.05518666572040982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,5120,0.09361688958273993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,6144,0.04766399992836846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,7168,0.14901688363817003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,4096,0.07209333446290758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,6144,0.13362488481733534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,5120,0.04152977797720168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,5120,0.1027546657456292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,4096,0.033879998657438494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,4096,0.08198755317264132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,16384,0.12642844518025717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,3584,0.03086400032043457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,2560,0.04450222187572055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,3584,0.07233244180679321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,3072,0.02765422231621212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,2048,0.03750844465361701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,2560,0.0243493335114585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,3072,0.06343377961052789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,10240,0.07515733109580146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,1536,0.026543999711672466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,2560,0.054620444774627686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,1024,0.01925511161486308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,1536,0.01815022197034624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,2048,0.04637777805328369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,768,0.016128000285890367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,1536,0.03695999913745456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,1024,0.01551555593808492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,1024,0.027504000398847792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,512,0.012216889195972018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,768,0.014472888575659858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,256,0.008779555559158325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,768,0.023520888553725347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,512,0.013524444566832649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,3584,0.060927112897237144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,128,0.007251555720965068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,512,0.01903377804491255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,256,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,64,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,3072,0.05222310953670078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,32,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,128,0.012411555482281579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,256,0.016490666402710807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,128,0.015804444750150044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,64,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,32,0.015843555331230164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,2048,0.02122488949033949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,65536,0.4348728921678331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,16384,0.2503768867916531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,65536,0.8630373213026258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,16384,0.10299288564258152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,12288,0.19835733042822945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,51200,0.7790995703803169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,16384,0.3003244400024414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,12288,0.07829688655005561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,10240,0.181931561893887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,12288,0.23714754316541883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,10240,0.06884621911578707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,8192,0.13743555545806885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,51200,1.1155199474758573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,7168,0.12597510549757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,8192,0.05566044648488363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,65536,1.4988266626993816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,6144,0.1082035568025377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,8192,0.18515200085110137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,7168,0.05297955539491442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,7168,0.12518133057488337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,6144,0.0470382206969791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,4096,0.06184266673194038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,6144,0.10890311002731323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,5120,0.03787911269399855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,3584,0.061257779598236084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,51200,0.3400159941779242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,4096,0.03052622079849243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,3072,0.050316443045934044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,4096,0.07799733347362943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,3584,0.0280693338976966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,3584,0.0658586687511868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,3072,0.0255084451701906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,2048,0.03305955396758185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,3072,0.05806666612625122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,2560,0.022652443912294176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,1536,0.026570666167471144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,2560,0.050329777929517955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,2048,0.019895111521085102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,1024,0.01649599936273363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,1536,0.01754577789041731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,10240,0.23000621795654297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,768,0.013185777597957186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,1536,0.03401866555213928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,1024,0.01481066644191742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,512,0.010806221928861407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,5120,0.08957688675986396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,1024,0.025285333395004272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,256,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,768,0.02145244512293074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,512,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,5120,0.09658310810724895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,128,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,256,0.012062222593360476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,64,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,256,0.015185778339703878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,2560,0.04338044590420193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,128,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,32,0.006220444622966979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,128,0.014537778165605334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,64,0.014487110906177096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,32,0.0147733340660731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,2048,0.0421982208887736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,65536,0.38554400867886013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,768,0.013770666387346057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,65536,0.7614133093092176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,512,0.01754666699303521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,51200,0.7429928779602051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,16384,0.227147552702162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,51200,0.30667111608717174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,16384,0.09425777859157985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,12288,0.20190933015611437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,16384,0.26902222633361816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,10240,0.14831822448306614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,12288,0.07350489166047838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,12288,0.1975395547019111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,51200,0.8995911280314127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,8192,0.11616710821787517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,10240,0.06101422177420723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,7168,0.09864266713460286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,8192,0.04743911160363091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,6144,0.08802222543292576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,10240,0.20705155531565347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,7168,0.04264977905485365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,5120,0.06967555814319186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,8192,0.1284044451183743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,4096,0.054562668005625405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,6144,0.037314666642083064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,5120,0.032491554816563926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,3584,0.0454346670044793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,6144,0.09796000189251369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,3072,0.039876444472206965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,5120,0.08666311370001899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,3584,0.024898666474554274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,4096,0.06566666894488864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,2560,0.033422221740086876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,3072,0.023235556152131822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,3584,0.06291911337110731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,2048,0.027514666318893433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,3072,0.05200622147983975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,1536,0.02092711130777995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,2560,0.02052088909678989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,2048,0.018219555417696636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,2560,0.044362667534086436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,65536,1.1714080174763997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,1024,0.014800000521871777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,2048,0.03808622227774726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,768,0.012207110722859701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,1536,0.030287110143237646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,512,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,1024,0.02257333364751604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,768,0.019193778435389202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,256,0.006872889068391588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,512,0.011811555259757571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,512,0.015776889191733468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,128,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,256,0.011381333072980246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,256,0.013636444177892474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,4096,0.02721422248416477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,64,0.005543111099137201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,32,0.005538666827811136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,128,0.011083555718262991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,128,0.01314755529165268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,64,0.01311911145846049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,32,0.013405332962671915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,1536,0.015591111448076038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,1024,0.013634666800498962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,768,0.012891555825869242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,65536,0.316922664642334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,65536,0.6942755381266276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,51200,0.4376960065629747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,16384,0.18010666635301378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,12288,0.13306311766306558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,7168,0.11584266026814778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,16384,0.09675644503699408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,10240,0.11246755388047959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,12288,0.06782844331529406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,65536,0.9593555662367078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,16384,0.24525154961480033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,8192,0.09072355429331462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,10240,0.05297155512703789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,7168,0.07826755444208781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,8192,0.04145777887768216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,51200,0.7470533582899305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,10240,0.1633653375837538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,6144,0.06747111346986559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,8192,0.11694844563802083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,7168,0.03910933269394769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,5120,0.05384000142415365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,6144,0.03355822298261855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,4096,0.04498222139146593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,7168,0.09796000189251369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,6144,0.08893333541022407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,4096,0.02474311159716712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,5120,0.07177422444025676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,3072,0.03596444593535529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,4096,0.05710311068428887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,3584,0.022006221943431433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,3584,0.05367733372582329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,2560,0.030557334423065186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,51200,0.25115378697713214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,3072,0.04570755693647596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,2048,0.02546666728125678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,2560,0.01831733352608151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,12288,0.16184177663591173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,1536,0.019887111253208585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,2048,0.016228444046444364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,2048,0.033428443802727595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,1536,0.014501333236694336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,1536,0.027278222971492346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,5120,0.029656890365812514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,768,0.012430222498046027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,1024,0.02021422154373593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,768,0.011737777955002256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,512,0.010001777774757808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,3072,0.02014222244421641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,512,0.011008888483047485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,512,0.014127110441525778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,2560,0.039706667264302574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,128,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,256,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,256,0.012443555725945367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,1024,0.014415999253590902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,64,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,128,0.011752000285519494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,32,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,3584,0.04002933369742499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,1024,0.012676444318559436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,64,0.011716444459226398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,32,0.012040889097584618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,768,0.0169831116994222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,65536,0.5429795583089193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,65536,0.2894879976908366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,51200,0.4565413263108995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,16384,0.15602044264475504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,256,0.007527111305130853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,51200,0.22958845562405059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,16384,0.07362933291329278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,65536,0.7682044241163465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,51200,0.5998835563659668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,10240,0.1011600030793084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,12288,0.05840622054206001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,8192,0.08139644728766547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,12288,0.13958577315012613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,10240,0.050399998823801674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,7168,0.06507911284764607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,8192,0.04238311118549771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,10240,0.12112622790866429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,8192,0.09427555402119954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,6144,0.05897421969307793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,7168,0.0344142218430837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,5120,0.04598044355710348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,7168,0.08336177799436782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,6144,0.03106844425201416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,4096,0.03690133492151896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,5120,0.027963555521435205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,3584,0.030699554416868422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,5120,0.060736000537872314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,4096,0.022826666633288067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,3072,0.02772444486618042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,4096,0.05024977856212192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,3584,0.020800888538360596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,12288,0.11589688724941677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,2560,0.022075555390781824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,3072,0.01920622256067064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,3584,0.04502933224042257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,2048,0.01908266709910499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,16384,0.18551554944780138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,3072,0.03996088769700792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,2560,0.01752711170249515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,1536,0.01552977826860216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,2048,0.015697777271270752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,1024,0.010781333678298526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,128,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,2560,0.03431199987729391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,768,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,1536,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,2048,0.028947555356555518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,512,0.007093333535724216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,1024,0.01204711116022534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,1536,0.023540443844265405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,256,0.005535111245181825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,768,0.011408888631396823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,1024,0.017866666118303936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,128,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,768,0.015154666370815702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,64,0.004257777912749184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,512,0.012618666721714867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,128,0.010096888575288985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,32,0.004528000122971005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,256,0.01091288857989841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,128,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,64,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,32,0.010370666782061258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,65536,0.24117689662509492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,51200,0.33556445439656574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,51200,0.19261156188117134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,51200,0.48097780015733504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,16384,0.11434133847554524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,512,0.010605333579911126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,256,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,12288,0.08643200000127156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,16384,0.06604888704088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,6144,0.07316888703240289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,10240,0.0764791104528639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,16384,0.15312443839179143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,12288,0.048927999205059476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,8192,0.06102399693595039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,12288,0.11468977398342556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,10240,0.043030222256978355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,7168,0.05375644564628601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,8192,0.03270666797955831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,10240,0.09733777576022679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,65536,0.46990399890475804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,6144,0.04448533389303419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,5120,0.036658667855792575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,7168,0.029304001066419814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,7168,0.06927111413743761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,6144,0.02603111167748769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,4096,0.029653333955340918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,5120,0.022774222824308608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,6144,0.060601777500576444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,65536,0.6098417705959743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,3584,0.02605244517326355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,4096,0.019652444455358718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,5120,0.050660444630516895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,3584,0.018039999736679923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,2560,0.019191111127535503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,3072,0.01661688917213016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,3072,0.03380889031622145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,2048,0.01584533353646596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,2560,0.015421332584487068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,2560,0.02900622288386027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,1536,0.012895999683274163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,2048,0.014141332772043018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,2048,0.024791111548741657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,1024,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,1536,0.01277955538696713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,8192,0.07829688655005561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,768,0.00794666674402025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,1536,0.02011644509103563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,1024,0.011057777537239922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,512,0.006320000108745363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,1024,0.015325332681337992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,768,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,256,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,768,0.01313511116637124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,512,0.00997155573632982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,3072,0.022376888328128394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,128,0.004547555413511065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,512,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,4096,0.0418177776866489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,256,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,128,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,32,0.004204444587230682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,256,0.00965244405799442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,128,0.009282666775915358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,64,0.009023110899660323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,32,0.008967999782827165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,65536,0.312370671166314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,51200,0.2406568792131212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,16384,0.07766844166649713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,51200,0.1497724453608195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,12288,0.057096891933017306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,16384,0.12058488527933757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,51200,0.3704480065239801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,10240,0.048635555638207324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,12288,0.04125244418780009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,12288,0.09139466947979397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,8192,0.038318223423428006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,64,0.003867555409669876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,10240,0.03675999906327989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,10240,0.07675200038486056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,7168,0.0341288877858056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,3584,0.03761955433421665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,6144,0.0270453327231937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,7168,0.024100444383091394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,8192,0.06203555398517185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,5120,0.023637332850032385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,7168,0.05463288889990913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,65536,0.18746756182776558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,6144,0.04821688930193583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,4096,0.018551111221313477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,5120,0.018936889039145578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,3584,0.017017778423097398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,5120,0.04070399867163764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,4096,0.016547555724779766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,16384,0.052858667241202466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,4096,0.03382222188843621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,3584,0.015408888459205627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,2560,0.012906666431162091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,3584,0.030216889248953924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,3072,0.014310222533014087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,3072,0.027239999837345544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,2560,0.013110222087966071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,2560,0.023363555471102398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,2048,0.012103111379676394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,1536,0.008729777402347988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,2048,0.020104888412687514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,8192,0.02649066680007511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,1024,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,1536,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,1536,0.01657511128319634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,6144,0.021765333082940843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,768,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,1024,0.00999644481473499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,1024,0.012802666260136498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,512,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,768,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,256,0.0038817777401871155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,65536,0.46419821845160586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,512,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,128,0.0034924443397257063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,512,0.009535111486911774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,256,0.008642666869693333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,64,0.0031831111345026228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,256,0.00813422186507119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,128,0.008662222160233391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,32,0.0031884445084465873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,2048,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,128,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,65536,0.22048710452185738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,32,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,65536,0.18578933344946968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,51200,0.17152533266279432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,16384,0.06450488832261828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,65536,0.41626400417751735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,51200,0.1497324440214369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,12288,0.049267556932237416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,51200,0.3246986601087782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,16384,0.05094844433996412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,3072,0.014888000157144336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,768,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,12288,0.040376888381110296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,8192,0.03176088796721564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,10240,0.03511910968356662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,12288,0.08198933468924628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,7168,0.026968000663651362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,10240,0.06902399990293714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,8192,0.02648711038960351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,6144,0.02457688914404975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,64,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,7168,0.023704888092146978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,5120,0.020668443706300523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,6144,0.021213332811991375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,6144,0.0433555543422699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,4096,0.016591999265882704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,5120,0.018600889378123812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,3584,0.015325332681337992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,5120,0.03700000047683716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,4096,0.016399999459584553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,3072,0.01386666629049513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,4096,0.03110577662785848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,3584,0.01554666625128852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,2560,0.012147555748621622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,10240,0.041145778364605375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,3584,0.02751111156410641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,2048,0.010216000179449717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,3072,0.014097778333557976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,3072,0.024452444579866197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,2560,0.013136000268989138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,1536,0.008259555531872643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,2048,0.012079111403889127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,2560,0.02143999934196472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,1024,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,2048,0.01828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,1536,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,1024,0.009923555784755284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,1536,0.01516266663869222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,8192,0.055416001213921436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,512,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,1024,0.011863110793961419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,7168,0.04945955673853556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,768,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,256,0.003912888881233004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,128,0.0034106667670938703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,768,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,512,0.009023110899660323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,64,0.003170666595300039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,256,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,512,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,32,0.0031902222997612427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,128,0.008519111408127679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,256,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,128,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,65536,0.18784711096021864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,64,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,32,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,51200,0.13914933469560412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,16384,0.11652266979217529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,65536,0.1401075522104899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,16384,0.046703110138575234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,12288,0.034155554241604276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,51200,0.11243911584218343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,16384,0.03962399893336826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,65536,0.3372008800506592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,10240,0.02696977721320258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,12288,0.03110044532352024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,51200,0.2643928792741564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,12288,0.06733955277336968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,768,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,7168,0.018569777409235638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,10240,0.0564791096581353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,8192,0.019348444210158456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,6144,0.016572443975342643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,8192,0.04617422156863742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,7168,0.017672888106769986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,5120,0.014557333456145393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,7168,0.04115733173158433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,6144,0.01625155574745602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,4096,0.012288000020715924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,6144,0.03590044379234314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,5120,0.014441778262456259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,3584,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,5120,0.030785779158274334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,4096,0.01277599980433782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,4096,0.025755556093321905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,3584,0.012103111379676394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,3584,0.02292355563905504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,2560,0.00868622213602066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,3072,0.011417778001891242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,3072,0.020615999897321064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,16384,0.08898666832182144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,2048,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,8192,0.021215111017227173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,10240,0.02722399930159251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,2560,0.011060444845093621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,1536,0.005704000178310607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,1024,0.004514666481150521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,2560,0.017913776967260573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,2048,0.010031110710567897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,768,0.0038782221575578055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,1536,0.009549332989586724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,2048,0.015557333827018738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,1024,0.008961777720186446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,512,0.0034977777136696708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,1536,0.013035555680592855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,768,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,1024,0.010332444475756751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,768,0.00907466643386417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,512,0.00812711069981257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,128,0.0027386666172080566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,256,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,512,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,64,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,32,0.002625777696569761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,256,0.006679111056857639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,128,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,3072,0.009970666633711921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,65536,0.10672889153162639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,64,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,51200,0.08532177739673191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,32,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,16384,0.03193511234389411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,65536,0.13819911744859484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,65536,0.295031123691135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,51200,0.11091378000047471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,12288,0.02212711175282796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,10240,0.016923555069499545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,16384,0.03885244329770406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,51200,0.23148621453179252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,16384,0.07693333095974393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,12288,0.030586666531032983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,10240,0.02681066592534383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,12288,0.05828977955712212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,7168,0.012301333248615265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,256,0.003162666741344664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,8192,0.019088000059127808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,6144,0.010947555303573608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,8192,0.040138665172788836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,7168,0.01758399936887953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,5120,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,6144,0.016195555528004963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,6144,0.031791110833485924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,5120,0.014356444279352823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,128,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,3584,0.007344000041484833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,5120,0.026692445079485577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,4096,0.012726222475369772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,3072,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,4096,0.022534221410751343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,3584,0.012103999654452005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,2560,0.005629333357016246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,3584,0.02035466664367252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,3072,0.011412444214026133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,2048,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,3072,0.018233777748213876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,2560,0.010804444551467896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,8192,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,1536,0.00452533322903845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,2048,0.010308444499969482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,1024,0.0035333335399627686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,2048,0.01408622165520986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,1536,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,10240,0.04924622178077698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,768,0.0032275555034478507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,1536,0.011713777979214987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,1024,0.008800889054934183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,512,0.0028764444092909494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,1024,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,7168,0.03588355580965678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,768,0.00831911133395301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,4096,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,256,0.002566222308410539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,128,0.002161777681774563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,512,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,512,0.0074275558193524676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,256,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,64,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,32,0.0022248888595236675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,128,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,256,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,128,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,65536,0.07696977588865492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,64,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,51200,0.06225066714816623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,32,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,16384,0.02425333360830943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,65536,0.137990223036872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,2560,0.01620444489849938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,65536,0.25291911760965985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,51200,0.11076977517869736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,12288,0.015506666567590503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,16384,0.04006577862633599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,10240,0.013482666677898832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,51200,0.19839466942681205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,8192,0.012807110945383707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,12288,0.030836443106333416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,10240,0.02771111163828108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,12288,0.05084800057941013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,7168,0.011424889167149862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,768,0.00851288851764467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,8192,0.019660444723235238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,10240,0.0429351098007626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,6144,0.009771555662155151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,5120,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,7168,0.018518222702874076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,6144,0.01682400041156345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,7168,0.0310791101720598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,4096,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,5120,0.015159999330838522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,6144,0.02730133301681942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,3584,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,5120,0.02344622214635213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,4096,0.013463111387358772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,3072,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,4096,0.019928000039524503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,3584,0.012598222328556908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,2560,0.005199999858935674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,3584,0.01807466646035512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,3072,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,2048,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,2560,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,1536,0.0038942222793896994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,2560,0.014511111709806653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,1024,0.0032471112079090546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,16384,0.06651822063657972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,1536,0.009677333136399588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,768,0.003145777723855443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,1024,0.008902221918106079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,1024,0.008910222185982598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,512,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,768,0.008627555436558193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,768,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,256,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,512,0.007970666719807519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,512,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,128,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,256,0.007687110867765214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,256,0.005664000080691443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,64,0.0021226666867733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,128,0.0076248885856734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,32,0.0022888889329300984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,128,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,3072,0.016377778516875375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,65536,0.07098488675223456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,2048,0.010334221853150262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,51200,0.056636446052127414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,2048,0.01272266689274046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,32,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,12288,0.013473778135246701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,1536,0.01071111112833023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,10240,0.011723555624485016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,65536,0.25291289223564994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,8192,0.0350195566813151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,16384,0.06611644559436373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,8192,0.010235555469989777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,12288,0.05063377817471822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,7168,0.009420444567998251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,6144,0.009679999616410997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,5120,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,10240,0.042709334029091724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,4096,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,8192,0.03472177849875556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,7168,0.03087555699878269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,3584,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,6144,0.027049778236283198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,3072,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,5120,0.023476445012622412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,2560,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,4096,0.01994044416480594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,2048,0.004552000098758274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,3584,0.01814399990770552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,1536,0.003908444609906939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,3072,0.016200888488027785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,64,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,1024,0.003208000005947219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,768,0.002874666617976295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,2560,0.014465777410401238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,16384,0.021691555778185528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,512,0.002536000063021978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,2048,0.012531555361217923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,256,0.0024844445288181305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,128,0.0021893332401911416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,1536,0.01073688848151101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,64,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,1024,0.008823999928103553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,768,0.007878222399287755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,32,0.002190222135848469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,512,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,51200,0.19813688596089682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,256,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,128,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,65536,0.07250844770007663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,51200,0.056291553709242076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,16384,0.02057066725360023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,12288,0.011400889191362592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,64,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,10240,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,32,0.005158222383923001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,8192,0.008976000050703684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,7168,0.008868444297048781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,65536,0.2519822120666504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,16384,0.06618133518430921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,6144,0.008434666527642144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,10240,0.04282755653063456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,5120,0.007522666619883643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,8192,0.03460977805985345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,4096,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,7168,0.030844443374209937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,6144,0.02700177828470866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,3072,0.007383999725182851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,5120,0.023373333944214716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,4096,0.019792889555295307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,2560,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,2048,0.004518222063779831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,3072,0.016211556063758004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,1536,0.0038488888078265717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,2560,0.014476444986131458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,1024,0.0031679999083280563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,2048,0.012798222402731577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,768,0.0028364445186323593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,1536,0.010686222049925061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,512,0.002542222125662698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,256,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,768,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,51200,0.19763378302256265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,12288,0.0506933331489563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,64,0.0018835556175973679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,512,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,3584,0.008014221986134848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,32,0.0018800000349680583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,256,0.005635555419656966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,128,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,64,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,32,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,1024,0.008630221916569604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,128,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,3584,0.018214222457673814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,16384,1.2962835099962022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,16384,4.306385040283203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,16384,4.902577718098958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,51200,5.14500003390842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,12288,3.222260369194878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,16384,4.11586676703559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,12288,0.9335732989841037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,12288,3.2974781460232205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,10240,0.8241635428534614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,10240,2.7743973202175565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,10240,2.788982179429796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,12288,3.051367017957899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,8192,0.6698444684346517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,10240,2.51419915093316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,8192,2.621944851345486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,8192,2.1135760413275824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,51200,12.860767788357206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,8192,2.04640981886122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,7168,0.5964701970418295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,7168,2.062781227959527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,7168,1.9116800096299913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,6144,0.5269173516167535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,6144,1.7818035549587672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,7168,1.800871107313368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,6144,1.6142329110039606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,5120,1.434893290201823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,6144,1.5360541873508031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,5120,0.4393022325303819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,4096,1.0204231474134657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,5120,1.3525217903984919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,4096,0.36625689930386013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,4096,1.0828035142686632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,3584,0.8714115354749891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,5120,1.2121342553032768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,4096,0.9888488981458875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,3584,0.33177510897318524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,3584,0.9356418185763888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,3072,0.9010915756225586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,3072,0.791992876264784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,3584,0.8507724338107638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,3072,0.2915395630730523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,2560,0.2538364463382297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,2560,0.7697573237948947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,3072,0.7566746605767144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,51200,13.54468960232205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,2560,0.6567964553833008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,2048,0.46839019987318253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,2560,0.6132888793945312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,2048,0.22641865412394205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,2048,0.5476453569200304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,1536,0.1772079997592502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,1536,0.4266071054670546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,1536,0.4269146654340956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,2048,0.502928892771403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,768,0.20756622155507407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,1024,0.15384356180826822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,1024,0.3308302296532525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,51200,20.231050279405384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,768,0.13986666997273764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,768,0.26484089427524143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,512,0.15639821688334146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,1024,0.29911822742886013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,256,0.10540977451536392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,512,0.10787110858493382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,768,0.244704008102417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,128,0.09020177523295085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,1024,0.26954399214850533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,256,0.09320355786217584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,512,0.2214204470316569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,64,0.07640710804197523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,256,0.1580666700998942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,128,0.09128444724612766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,32,0.07937777704662748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,128,0.18346222241719565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,1536,0.39946667353312176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,64,0.18791288799709746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,32,0.18924089272816977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,256,0.19425600104861787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,512,0.19708800315856934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,128,0.1165884468290541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,51200,4.052259657118055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,65536,5.286628299289279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,51200,10.136009216308594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,16384,3.8883340623643665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,16384,3.942960951063368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,16384,1.0742791493733723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,65536,13.098815070258247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,12288,2.647229300604926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,12288,0.7375084559122721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,16384,3.2655084398057728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,12288,2.8768249087863498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,51200,10.954705132378471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,12288,2.3686161041259766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,10240,0.6498755349053277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,51200,16.395923190646702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,10240,2.4927919175889754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,10240,2.2404026455349393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,10240,2.0387545691596136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,8192,0.6109093560112847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,8192,1.932288063897027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,8192,1.8234195709228516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,65536,19.21923828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,7168,1.5158462524414062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,8192,1.5953973134358723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,7168,0.467499574025472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,7168,1.5095733006795247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,6144,0.39635999997456867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,65536,14.78104994032118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,6144,1.2387510935465496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,7168,1.3728986316257055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,5120,1.2319306267632377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,5120,0.4190977679358588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,6144,1.2150488959418404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,4096,0.8263315624660916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,6144,1.1703093846638997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,5120,1.0496844185723198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,4096,0.28577955563863117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,5120,0.9693653318617078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,4096,0.8245555559794108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,3584,0.7422453562418619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,3584,0.25904443528917104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,3072,0.64028443230523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,3072,0.2347573439280192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,4096,0.8178017934163412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,2560,0.5351306597391764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,3584,0.6472551027933756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,3072,0.641357315911187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,3072,0.5664426485697428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,2560,0.20360533396402994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,2048,0.44008710649278426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,2560,0.6391182475619847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,2048,0.16986221737331816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,2560,0.49475468529595273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,2048,0.4452240202162001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,1536,0.32996800210740834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,1024,0.2072977754804823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,1536,0.14115733570522732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,2048,0.396126217312283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,1536,0.3358453379737006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,1024,0.10836888684166802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,768,0.16608444849650064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,1536,0.319160885281033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,1024,0.30514756838480633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,768,0.10376977920532227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,512,0.13442577256096735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,1024,0.23933511310153535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,256,0.08791555298699273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,512,0.0866053303082784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,768,0.19891822338104248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,512,0.17399199803670248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,768,0.1945404476589627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,128,0.07135644223954943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,256,0.07483733362621732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,512,0.1556346681382921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,256,0.15377244684431288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,3584,0.7220266660054525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,64,0.058656891187032066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,128,0.14475999938117132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,32,0.0606844425201416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,128,0.0727893312772115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,256,0.1221804486380683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,128,0.09328266647126938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,64,0.14711911148495144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,32,0.1503031121359931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,51200,1.2071644465128581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,65536,1.6278808381822374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,16384,1.2180212868584526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,51200,3.2182114919026694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,16384,1.3083520465426974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,16384,0.3380888832939996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,65536,4.296199798583984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,12288,0.7841502295600044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,16384,1.0311217837863498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,12288,0.25949954986572266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,51200,3.5521519978841147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,12288,0.8875422477722168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,12288,0.7634000248379178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,51200,5.426194508870442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,10240,0.2207111120223999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,10240,0.6983057657877604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,10240,0.7122809092203776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,8192,0.5650577545166016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,8192,0.18507556120554605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,10240,0.6420693397521973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,65536,4.668501536051433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,8192,0.5747137599521214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,65536,6.576253255208333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,8192,0.4850631290011936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,7168,0.5360417895846897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,7168,0.1574675507015652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,6144,0.13930133978525797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,6144,0.435531563229031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,7168,0.5579608811272515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,6144,0.40756800439622665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,5120,0.4220160113440619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,5120,0.1336817741394043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,5120,0.361586676703559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,6144,0.3690764374203152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,4096,0.26819467544555664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,4096,0.1055484414100647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,3584,0.22637687789069283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,4096,0.2694942156473796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,5120,0.31046756108601886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,3072,0.19188355074988472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,3584,0.09576711389753555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,4096,0.26436087820265025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,3584,0.24663021829393175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,2560,0.1716151105033027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,3584,0.22544799910651314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,3072,0.2051102187898424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,2048,0.13951199584537083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,3072,0.20914578437805176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,2560,0.06931822167502509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,7168,0.4260853396521674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,1536,0.10354222191704644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,2560,0.1671466694937812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,2048,0.14388622177971735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,2048,0.059451553556654185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,1024,0.06922844383451673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,1536,0.049048887358771436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,1536,0.11427199840545654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,2048,0.14107288254631892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,768,0.05267911156018575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,1024,0.08782222535875107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,1024,0.040630221366882324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,1536,0.11176178190443252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,512,0.044623110029432506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,768,0.037300444311565824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,1024,0.08733866612116496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,768,0.0695448915163676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,256,0.028359999259312946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,3072,0.08678311109542847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,512,0.034352888663609825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,512,0.05957421991560194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,768,0.07390666670269437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,128,0.025919111238585577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,2560,0.1778435574637519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,256,0.030607998371124268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,256,0.05335999859703911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,512,0.05963022179073758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,64,0.02130222154988183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,32,0.021677333447668288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,128,0.050704889827304415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,128,0.029871109459135268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,256,0.04758844441837735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,64,0.05144355694452921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,32,0.052001777622434825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,128,0.036329779360029436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,65536,1.2166755464341905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,16384,0.705952008565267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,51200,2.679412417941623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,51200,1.0119413799709744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,65536,3.2550862630208335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,16384,0.30771199862162274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,16384,0.8905671437581381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,12288,0.5939208666483561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,51200,2.708546744452583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,51200,3.8673138088650174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,16384,0.7543804380628797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,12288,0.20156710677676729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,12288,0.6472222010294596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,10240,0.5433768696255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,10240,0.16944889227549234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,12288,0.5487768914964464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,65536,4.750234815809462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,65536,3.533507453070747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,10240,0.5447280142042372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,8192,0.40726399421691895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,8192,0.1356711122724745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,7168,0.3776560094621446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,7168,0.1240337822172377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,6144,0.2862222194671631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,8192,0.609023994869656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,8192,0.36588888698154026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,7168,0.45360178417629665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,6144,0.11033955547544692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,7168,0.3254942099253337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,6144,0.3701395458645291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,5120,0.317251549826728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,4096,0.18667288621266684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,5120,0.26030490133497447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,6144,0.2924133406745063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,4096,0.07705510987175836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,3584,0.16570044888390437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,4096,0.21511110994550917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,5120,0.24016088909573027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,3072,0.14297600587209067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,3584,0.06850488980611165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,10240,0.457747565375434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,4096,0.203076442082723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,3584,0.18411644299825033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,2560,0.13365511099497476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,3072,0.060940444469451904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,3072,0.1568924453523424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,3584,0.17550399568345812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,2048,0.10449688964419895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,2560,0.0539262228541904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,2560,0.1374515559938219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,1536,0.07980355289247301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,3072,0.15221244759029812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,2048,0.046887109676996865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,2560,0.12968356079525417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,1024,0.05764444669087728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,2048,0.1111146608988444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,5120,0.0919111106130812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,1536,0.03967644439803229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,768,0.04559288753403557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,1536,0.08866488933563232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,1024,0.03256622287962172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,2048,0.10963377687666152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,512,0.031152000029881794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,1536,0.08893689182069565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,768,0.030449777841567993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,768,0.05463822351561653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,1024,0.06931822167502509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,256,0.021984888447655573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,512,0.02805688977241516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,512,0.04629066586494446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,768,0.05853955613242256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,256,0.025123554799291823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,256,0.04155022237035964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,64,0.016509332590632968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,128,0.024691555235120986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,128,0.0395084453953637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,256,0.0384844442208608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,128,0.02978133492999607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,64,0.0400622222158644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,32,0.040468444426854454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,1024,0.06796444124645658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,128,0.01963733302222358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,512,0.04896533489227295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,32,0.01737333337465922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,65536,0.9667439990573459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,16384,0.620182196299235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,65536,2.7428586747911243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,51200,0.8746470875210233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,51200,2.2732160356309676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,16384,0.23636534478929308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,16384,0.763919088575575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,12288,0.5458613501654731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,16384,0.6615742047627767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,65536,3.944813198513455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,51200,3.0676212310791016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,12288,0.18136800660027397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,65536,2.991511027018229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,12288,0.6518035464816624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,10240,0.14172266589270696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,51200,2.3235768212212458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,12288,0.4749644597371419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,10240,0.4611475732591417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,8192,0.3808435599009196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,8192,0.11578310860527886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,7168,0.3003573417663574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,8192,0.35276177194383407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,10240,0.40534043312072754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,7168,0.1036764449543423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,6144,0.2689475483364529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,8192,0.3239111105600993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,7168,0.3131013446384006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,6144,0.0915679997868008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,5120,0.2346177895863851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,6144,0.26073066393534344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,7168,0.30541955100165474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,5120,0.07881689071655273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,4096,0.18584356043073866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,10240,0.4869190851847331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,5120,0.21696000629001191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,6144,0.2715120050642225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,3584,0.15175910790761313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,4096,0.1744204494688246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,5120,0.20843377378251818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,3584,0.0597591135236952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,3072,0.13081599606408015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,3584,0.15350488821665445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,2560,0.11320800251430935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,4096,0.17111733224656847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,3584,0.15043289131588405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,3072,0.1337840027279324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,2048,0.08984088897705078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,2560,0.046448889705869884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,2560,0.11574044492509629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,3072,0.1348151101006402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,1536,0.06181777848137749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,2048,0.04029066695107354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,2048,0.09476177559958564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,2560,0.11550222502814399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,1024,0.04785955614513821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,1536,0.03380533390574985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,1536,0.0758204460144043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,2048,0.0962631106376648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,768,0.03898222247759501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,1024,0.027794665760464136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,1024,0.05824977821773953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,4096,0.06594755252202351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,1536,0.07987733019722833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,512,0.02899644441074795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,768,0.025535111625989277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,768,0.046145776907602944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,1024,0.061055110560523145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,256,0.026111111044883728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,512,0.023522666758961145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,128,0.017700443665186565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,3072,0.05288355549176534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,512,0.038768887519836426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,768,0.052411556243896484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,256,0.02088266611099243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,64,0.014828445182906257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,32,0.014523555835088095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,512,0.043523556656307645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,256,0.03395288851526048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,128,0.02051911089155409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,128,0.03259199857711792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,256,0.034097777472601995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,64,0.03211466802491082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,32,0.03175910976197984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,128,0.02619733413060506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,65536,0.8143137825859917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,51200,0.6352124214172363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,16384,0.7872631284925672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,51200,1.8476888868543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,65536,2.162264082166884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,16384,0.8653448952568902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,16384,0.20270310507880318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,12288,0.4148755603366428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,51200,2.6394986046685114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,12288,0.14840622742970785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,12288,0.43529155519273544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,10240,0.3614906734890408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,12288,0.3999519877963596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,65536,3.382953855726454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,65536,2.493805355495877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,10240,0.11900799804263645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,10240,0.3775591055552165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,8192,0.2908408906724718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,8192,0.09642577833599514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,7168,0.2547279993693034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,8192,0.3569386535220676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,10240,0.46545865800645614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,8192,0.26891112327575684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,16384,0.5407350858052572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,7168,0.24977156851026747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,7168,0.08671022123760647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,6144,0.21976889504326713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,5120,0.17666133244832358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,6144,0.07585422197977702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,6144,0.21307110786437988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,7168,0.2663297653198242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,5120,0.06536799669265747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,5120,0.19864622751871744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,4096,0.13407556215922037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,6144,0.20305066638522676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,3584,0.11382488409678142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,4096,0.057229333453708224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,4096,0.14573600557115343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,5120,0.17343644301096597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,3072,0.09738044606314765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,3584,0.12624443901909724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,3584,0.049672888384925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,4096,0.14333955446879068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,2560,0.085844443904029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,3072,0.04377777708901299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,51200,1.9099191029866536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,3584,0.1410168939166599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,2048,0.06703466839260526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,2560,0.039396444956461586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,2560,0.09530399905310737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,1536,0.051671998368369214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,3072,0.1128728919559055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,2048,0.033717334270477295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,2048,0.07881155278947619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,2560,0.09666311078601414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,1024,0.03633866707483927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,1536,0.027423999375767175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,768,0.02684088879161411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,1536,0.06254044506284925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,2048,0.08330488867229886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,1024,0.022912888063324824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,512,0.019827556278970506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,1024,0.04775999983151754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,1536,0.06752977768580119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,768,0.02126400007141961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,768,0.038789331912994385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,256,0.01511111193233066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,512,0.01941244469748603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,1024,0.05212622218661838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,128,0.013879111243618859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,512,0.031269331773122154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,768,0.045075555642445884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,256,0.01755466726091173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,64,0.011712888876597086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,256,0.02725244396262699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,512,0.037068443165885076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,32,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,128,0.018181333939234417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,3072,0.11074222458733453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,128,0.026416889495319788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,64,0.02641333308484819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,128,0.02200088898340861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,32,0.02650044361750285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,65536,0.6765377786424426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,256,0.02983022067281935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,51200,1.4724061754014757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,16384,0.45205068588256836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,51200,0.6110542085435655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,16384,0.20024800300598145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,16384,0.5395324495103624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,51200,1.6524559656778972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,51200,2.3504986233181424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,12288,0.4073324468400743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,16384,0.4709102312723796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,65536,2.9603936937120228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,12288,0.38205777274237734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,10240,0.3116515477498372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,65536,2.004977756076389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,10240,0.10516177945666844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,10240,0.3335431151919895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,12288,0.37520800696478945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,10240,0.2925635443793403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,8192,0.2509964572058784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,8192,0.23192532857259116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,7168,0.22861777411566842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,65536,2.093218697441949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,7168,0.07592888673146565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,6144,0.18309599823421904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,7168,0.2048471106423272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,5120,0.1506284475326538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,6144,0.06690755817625257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,8192,0.26767200893825954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,6144,0.1878817743725247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,6144,0.17921245098114014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,8192,0.08516977892981635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,5120,0.058711111545562744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,5120,0.15944711367289224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,4096,0.11999111705356175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,3584,0.09773244460423787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,4096,0.04867911007669237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,5120,0.15164355436960855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,4096,0.12775466177198622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,12288,0.13324088520473903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,4096,0.12565510802798802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,3584,0.1158248848385281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,3584,0.11215910646650527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,3072,0.03894044293297662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,3072,0.09833066993289524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,2560,0.07487644751866658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,2560,0.03458755546145969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,3072,0.09946044286092122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,2560,0.08578488561842178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,2048,0.05755199988683065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,7168,0.21145244439442953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,1536,0.043311112456851535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,2048,0.029551111989551123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,2560,0.08684711323844062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,2048,0.07024088833067152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,1024,0.029118223322762385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,1536,0.055952886740366616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,2048,0.07395378086302015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,3072,0.09397511349784003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,3584,0.043155554268095225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,768,0.025024889243973628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,1024,0.021171554923057556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,1536,0.05960888995064629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,512,0.017315554949972365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,768,0.03506044546763102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,768,0.01868977811601427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,1024,0.0469413333468967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,256,0.013166222307417126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,512,0.01715022159947289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,768,0.04039377636379666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,128,0.010801778071456485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,256,0.02460799945725335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,256,0.015821332732836407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,512,0.033979554971059166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,64,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,256,0.026856889327367146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,128,0.015944888194402058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,32,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,1536,0.025064888927671645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,128,0.019938665959570143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,32,0.023393778337372675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,1024,0.042585776911841504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,512,0.028049776951471966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,65536,0.7063102192348905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,128,0.0233297778500451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,64,0.02404977712366316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,65536,1.5990906821356878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,51200,1.2331199645996094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,16384,0.41795910729302305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,51200,0.5492249064975315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,16384,0.1625795496834649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,16384,0.4689155684577094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,12288,0.34105245272318524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,16384,0.4277848932478163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,65536,2.5562871297200522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,12288,0.12785332732730442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,65536,1.8964523739284938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,12288,0.3575724495781793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,51200,1.9790098402235243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,10240,0.2585795455508762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,12288,0.3105306625366211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,10240,0.1072933342721727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,8192,0.2275235652923584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,10240,0.3259680006239149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,8192,0.07695555686950684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,7168,0.2067191070980496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,8192,0.2548906538221571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,10240,0.26929601033528644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,7168,0.06889600223965116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,6144,0.17598844899071586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,7168,0.20492977566189238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,8192,0.2158097823460897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,5120,0.14549867312113443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,6144,0.06075466341442532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,6144,0.17136000262366402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,7168,0.1911635531319512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,4096,0.11336889531877305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,5120,0.06223289171854655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,5120,0.14311911000145808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,6144,0.16516710652245417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,3584,0.1063199970457289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,4096,0.04384266667895847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,4096,0.11293244361877441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,3072,0.08424088690016006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,5120,0.1393724415037367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,3584,0.04755377769470215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,3584,0.09979733493593003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,2560,0.06834755341211955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,4096,0.11539199617173936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,3072,0.0864880018764072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,3584,0.10399022367265488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,2048,0.057757331265343555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,51200,1.4710044860839844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,3072,0.09156711233986749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,2560,0.03137866655985514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,1536,0.04212622178925408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,2048,0.02686133318477207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,2560,0.07894488837983873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,2048,0.062071111467149526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,1024,0.03127822279930115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,1536,0.022341334157519873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,1536,0.04942489001486036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,2048,0.06643199920654297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,768,0.022636445032225713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,1024,0.018959111637539335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,1536,0.055439111259248525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,1024,0.037795556916130915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,512,0.01666755477587382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,768,0.017669333351982962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,3072,0.03460355599721273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,512,0.02496355606449975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,256,0.012079111403889127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,512,0.01586666703224182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,768,0.036632001399993896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,2560,0.07511555486255221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,128,0.009954666925801171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,256,0.014781332678265043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,512,0.03127022253142463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,256,0.02162222233083513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,64,0.008622222476535374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,128,0.0148044443792767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,128,0.020960888928837244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,32,0.009000000026490953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,256,0.023176888624827068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,64,0.020977778567208186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,32,0.020789333515697055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,1024,0.043113778034845986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,65536,0.5789368947347006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,65536,1.3366372850206163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,128,0.018473777506086562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,51200,1.1578906377156575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,768,0.03115733464558919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,16384,0.35616087913513184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,51200,0.44687912199232316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,65536,1.9047387440999348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,16384,0.13143732812669542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,51200,1.5449351204766169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,12288,0.2668462329440647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,16384,0.4268266624874539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,65536,1.6150248845418294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,12288,0.09742044740253025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,10240,0.25157067510816783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,12288,0.28934844334920246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,16384,0.38077423307630753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,10240,0.08314488993750678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,8192,0.20419377750820586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,10240,0.24126667446560332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,12288,0.29015644391377765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,51200,1.26274840037028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,8192,0.06731022066540189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,7168,0.17995733684963652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,8192,0.18642844094170463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,10240,0.25087377760145396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,6144,0.13071289327409533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,7168,0.06784177488750882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,7168,0.16326222154829237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,5120,0.09777066442701553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,6144,0.054336888922585376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,6144,0.16867644257015654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,4096,0.08992177910274929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,7168,0.16950755649142793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,5120,0.04509333438343472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,5120,0.12480533123016357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,3584,0.07349689139260186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,4096,0.03783644570244683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,4096,0.09977510902616714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,5120,0.12663288911183676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,3072,0.06248444318771362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,3584,0.0343466666009691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,3584,0.0872195561726888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,4096,0.10557688607109918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,2560,0.053314665953318276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,3072,0.031325333648257785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,3584,0.09257600042555068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,3072,0.07501422034369574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,2048,0.046748446093665234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,2560,0.06509511338339911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,3072,0.082896888256073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,1536,0.03560622202025519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,8192,0.1921537717183431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,2560,0.07241688834296332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,2048,0.05380622214741177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,1024,0.02624977793958452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,1536,0.02051466703414917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,2048,0.060487998856438525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,1536,0.04356266723738777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,768,0.01963555481698778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,6144,0.14814044369591609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,1024,0.03247288862864176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,1024,0.017895999881956313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,1536,0.0495324432849884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,512,0.015060444672902426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,768,0.027349332968393963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,768,0.01625422139962514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,1024,0.03938666648334927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,256,0.010649777948856354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,512,0.021935999393463135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,768,0.03269688950644599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,2560,0.02780710988574558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,256,0.019259555472267997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,128,0.009481777747472128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,256,0.013704000247849358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,2048,0.024136000209384497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,64,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,128,0.01386222243309021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,256,0.021850667066044275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,32,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,128,0.018260444204012554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,128,0.01716977854569753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,64,0.018214222457673814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,32,0.018193778064515855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,512,0.014723555909262763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,65536,0.4548853238423665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,512,0.027207111318906147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,65536,1.1339129341973198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,51200,0.8888186878628201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,16384,0.3108631239997016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,51200,0.3561440043979221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,65536,1.7267253663804796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,16384,0.3845084508260091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,65536,1.3094524807400174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,51200,1.198936038547092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,16384,0.3089919884999593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,12288,0.08311733272340563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,12288,0.2678977648417155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,10240,0.17224533028072783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,51200,1.03294128841824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,12288,0.23397954305013022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,8192,0.13162133428785536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,10240,0.06989066468344794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,7168,0.1311991082297431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,8192,0.06252266963322957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,8192,0.17155822118123373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,10240,0.19797689384884307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,6144,0.10696266757117377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,8192,0.15815644794040257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,12288,0.21173777845170763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,7168,0.13800533612569174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,5120,0.0859031081199646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,6144,0.04403377903832329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,7168,0.14381510681576198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,6144,0.11901333596971299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,4096,0.06606666909323798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,5120,0.03826844361093309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,5120,0.10059733523262872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,6144,0.12250310844845241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,3584,0.059057778782314725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,4096,0.03176444437768724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,5120,0.10649422142240737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,4096,0.08067555560006036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,10240,0.19215910964541963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,3072,0.04878933231035868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,3584,0.028956443071365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,3584,0.07127910852432251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,4096,0.08726844522688125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,2560,0.03965244359440274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,3072,0.02624711063173082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,7168,0.05148622062471178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,3072,0.0632657806078593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,3584,0.07874755726920234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,2048,0.03377866744995117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,2560,0.054571555720435254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,3072,0.06886844502554999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,1536,0.027627555860413447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,2048,0.02049422264099121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,2560,0.06022310919231839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,2048,0.04537422127193875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,1536,0.017870222528775532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,1536,0.03656444284651015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,2048,0.051514665285746254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,768,0.01533777846230401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,1024,0.01516888870133294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,1536,0.04195911023351881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,1024,0.027266666293144226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,512,0.011062222222487131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,768,0.023226666781637404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,768,0.014215999179416232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,512,0.013051555388503604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,512,0.018696889281272888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,256,0.008978666530715095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,768,0.02803200152185228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,16384,0.118487106429206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,128,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,2560,0.023702222439977858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,256,0.012170666621790992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,256,0.01624000072479248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,64,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,128,0.012167111039161682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,128,0.015522667103343539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,256,0.021021333005693223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,32,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,1024,0.01889600025282966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,128,0.01645511057641771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,64,0.015559110376569959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,32,0.015470221638679504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,1024,0.032772445016437106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,65536,0.4123031033409967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,65536,0.9722098244561089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,512,0.023593778411547344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,51200,0.6802275445726184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,16384,0.25358133845859104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,65536,1.414239141676161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,51200,0.33096443282233345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,16384,0.30057867368062335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,51200,1.0609635247124565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,12288,0.1976933346854316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,16384,0.10377955436706543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,10240,0.17758577399783662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,16384,0.27413421207004124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,12288,0.21067467000749376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,51200,0.8725235727098254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,10240,0.06640266709857516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,8192,0.13930400212605795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,12288,0.2081235514746772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,10240,0.18016444312201607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,7168,0.12398489316304524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,10240,0.1735724475648668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,8192,0.053298665417565234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,8192,0.14009155167473689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,6144,0.1010977758301629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,7168,0.046105778879589505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,7168,0.12347022692362468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,8192,0.15045777956644693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,65536,1.1746870676676433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,6144,0.10691289107004802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,7168,0.12516355514526367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,4096,0.06453600194719103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,5120,0.03460888730155097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,5120,0.08996088637246026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,5120,0.09505866633521186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,4096,0.02978844443957011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,4096,0.07230399714575873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,3584,0.0569653312365214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,4096,0.0785377754105462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,3584,0.026179555389616225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,3072,0.04900888933075798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,3584,0.0647653341293335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,3584,0.06991822189754911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,3072,0.023354666100607976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,5120,0.08016533321804471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,2560,0.04204622242185804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,3072,0.05724622143639458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,6144,0.04053777787420485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,2048,0.03454666667514377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,2560,0.02159022291501363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,3072,0.06272800101174249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,12288,0.07783377832836576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,1536,0.02475999957985348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,2560,0.049201776583989464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,6144,0.11071555482016669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,2048,0.018561777141359117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,2048,0.04126844472355313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,1024,0.016216889023780823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,2560,0.05387111173735725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,768,0.013958222336239286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,1536,0.01642400026321411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,2048,0.04706577791108025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,1024,0.014152000347773233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,1536,0.033026665449142456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,512,0.010738666686746808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,1024,0.024843555357721116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,768,0.013398222625255585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,256,0.007985778152942657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,1536,0.03826133410135905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,768,0.0210826661851671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,1024,0.029466665453381006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,512,0.012296888563368054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,128,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,512,0.01720177796151903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,256,0.011889778077602386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,768,0.025784000754356384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,64,0.006247111078765657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,256,0.015077332655588785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,512,0.021645334031846788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,128,0.011375111010339526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,128,0.01441422270403968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,256,0.019547555181715224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,128,0.014865777558750577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,32,0.014463999205165438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,65536,0.36388444900512695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,65536,0.8745750851101346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,51200,0.5712453524271647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,16384,0.21354399787055123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,65536,1.0783875783284504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,32,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,51200,0.8870159784952799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,51200,0.3028968969980876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,64,0.014440889159838358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,16384,0.08905866410997178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,12288,0.17125866148206922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,65536,1.1206364101833768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,16384,0.25699999597337514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,10240,0.1382062170240614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,12288,0.07020444340176053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,12288,0.21657866901821562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,16384,0.2733546627892388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,8192,0.11560711595747207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,10240,0.0616133345497979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,51200,0.8650995360480415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,10240,0.1626337766647339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,12288,0.2015226682027181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,7168,0.10023733642366196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,8192,0.04743999905056424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,6144,0.08646577596664429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,8192,0.12521689467959934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,10240,0.19032266404893664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,7168,0.05054577853944567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,7168,0.10929422246085273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,5120,0.06889688968658447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,8192,0.1404986646440294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,6144,0.039846221605936684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,4096,0.056088891294267446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,7168,0.1225351095199585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,5120,0.03207644489076402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,5120,0.08140177859200372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,6144,0.10816622442669338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,3584,0.04419111212094625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,4096,0.027727999620967444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,4096,0.06512888934877184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,3072,0.037943111525641546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,3584,0.02494311167134179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,4096,0.0758462217119005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,3584,0.0570933355225457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,2560,0.03156266609827677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,3072,0.022495999932289124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,3072,0.050007111496395536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,3584,0.06951110892825656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,2048,0.027254222167862788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,2560,0.020305777589480083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,2560,0.0433937774764167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,3072,0.060736000537872314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,1536,0.021260443660948012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,2048,0.017903111047214933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,2048,0.03661333190070258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,2560,0.05258488986227247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,1024,0.014840889308187695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,1536,0.015386665860811869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,6144,0.09629422426223755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,2048,0.04472800095876058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,768,0.012067555553383298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,1024,0.02239377796649933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,1536,0.037754668129815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,512,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,5120,0.09269689189063178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,768,0.012732444538010491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,1024,0.028802666399214003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,256,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,768,0.025382222400771245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,512,0.012044444680213928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,512,0.015466666883892484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,128,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,256,0.011067555182509951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,512,0.021459556288189355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,128,0.013087111214796701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,256,0.018239999810854595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,64,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,32,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,128,0.014511111709806653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,1536,0.029665778080622356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,64,0.01275022245115704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,32,0.01313511116637124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,1024,0.01346666696998808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,768,0.018992889258596633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,65536,0.31620799170600045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,65536,0.699645307328966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,51200,0.5324951277838813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,256,0.013444444371594323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,16384,0.17948799663119844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,128,0.011071110765139261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,51200,0.2843617863125271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,65536,0.9380506939358182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,16384,0.2116346624162462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,16384,0.07666044765048556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,51200,0.7426480187310113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,12288,0.1323644452624851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,65536,0.9592871136135526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,10240,0.12071288956536187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,12288,0.059804446167416044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,16384,0.2663991186353895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,8192,0.09766133626302083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,12288,0.16288710965050593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,10240,0.053787555959489614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,10240,0.1638773282368978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,7168,0.08586488829718696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,51200,0.7231458028157552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,12288,0.17811644077301025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,8192,0.040896889236238264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,6144,0.06167199876573351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,7168,0.04074933462672763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,8192,0.11258221997155084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,7168,0.10074488984213935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,8192,0.1230800019370185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,6144,0.03257333238919576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,5120,0.057835552427503795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,6144,0.08337510956658258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,7168,0.10874222384558783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,4096,0.044234666559431285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,5120,0.028301331732008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,5120,0.07084266344706218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,6144,0.0972391102049086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,3584,0.03892622060245938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,4096,0.024288889434602525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,4096,0.05703910854127672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,3072,0.03622577918900384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,5120,0.0817013316684299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,3584,0.02200266718864441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,4096,0.06751644611358643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,3584,0.051572445366117686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,2560,0.028967999749713477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,3072,0.020051555501090158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,3072,0.045238223340776235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,3584,0.06095200114780002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,2048,0.0243520008193122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,2560,0.017902221944597032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,2560,0.039232889811197914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,3072,0.054692443874147206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,1536,0.019890666007995605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,2048,0.016142222616407607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,2048,0.03321689036157396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,2560,0.04725689027044508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,1024,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,1536,0.02693422304259406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,1536,0.01443377799457974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,2048,0.03993777765168084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,768,0.01162755572133594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,1024,0.012681778106424542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,1024,0.020181333025296528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,1536,0.03357777661747403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,512,0.008943999807039896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,768,0.01702844434314304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,1024,0.026143999563323125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,256,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,512,0.014111111561457316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,512,0.01109155515829722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,768,0.022668444448047217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,128,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,256,0.010392000277837118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,512,0.019205333458052743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,256,0.012223111258612739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,64,0.005169777820507686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,10240,0.1511813269721137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,128,0.011715555356608497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,32,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,128,0.010461333725187512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,256,0.015861334072219003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,64,0.011695110963450538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,128,0.012923555241690742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,32,0.011700444751315646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,65536,0.28719109959072536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,65536,0.5608044200473362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,51200,0.373565329445733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,768,0.011720000041855706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,16384,0.1487520005967882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,65536,0.7627742025587293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,16384,0.093304885758294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,51200,0.5913155343797472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,12288,0.1038186682595147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,65536,0.7819555600484213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,16384,0.19693511062198213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,12288,0.05841866466734144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,51200,0.601035541958279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,10240,0.09251644213994344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,12288,0.16811645030975342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,8192,0.07020533084869385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,10240,0.05191644363933139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,12288,0.15231200059254965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,10240,0.11633955107794867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,7168,0.06591200166278414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,8192,0.03847022189034356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,8192,0.09585599766837226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,10240,0.12934133741590711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,6144,0.05702933337953356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,7168,0.03466755482885573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,8192,0.1045431097348531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,7168,0.08293777704238892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,5120,0.04407199886110094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,6144,0.030124445756276447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,51200,0.2263031005859375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,7168,0.09243200222651164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,16384,0.1825653314590454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,5120,0.026319111386934917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,5120,0.06064089139302572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,6144,0.07999910910924275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,3584,0.030857778257793848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,4096,0.02238222294383579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,4096,0.049439999792310924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,5120,0.06960888703664143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,3072,0.025384889708624944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,3584,0.01930488811598884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,3584,0.044475555419921875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,4096,0.0581208864847819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,2560,0.02288888891537984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,3584,0.0522017776966095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,3072,0.03955644369125366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,2560,0.03429333368937174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,3072,0.047028445535235934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,2048,0.018836443622907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,6144,0.07213955455356173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,2560,0.04137511054674784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,4096,0.03557777735922072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,2048,0.02898755669593811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,2048,0.015660444895426433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,2048,0.03491555651028951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,1536,0.012947555217478009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,1536,0.02325155503220028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,1536,0.02847733431392246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,1024,0.017571555243598092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,768,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,3072,0.01755555636352963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,1024,0.011352000137170156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,1024,0.02264444364441766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,512,0.006913777854707506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,768,0.01071200023094813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,2560,0.016656888855828177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,768,0.015189333094490899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,256,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,512,0.010050666828950247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,512,0.012446222205956778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,1536,0.015031110909250049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,128,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,256,0.01074044406414032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,256,0.009552000297440423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,512,0.0184551113181644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,64,0.00452533322903845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,1024,0.011362666885058085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,128,0.009646221995353699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,32,0.004573333180612988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,128,0.010395555860466426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,256,0.015549333559142219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,64,0.010330666270520952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,128,0.012423999607563019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,32,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,65536,0.44815376069810653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,65536,0.23453956180148652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,51200,0.30418488714430064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,16384,0.10843911435869004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,65536,0.6032711135016547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,51200,0.19712177912394205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,768,0.01996799972322252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,51200,0.47985421286688906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,16384,0.15191733837127686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,16384,0.061540444691975914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,65536,0.7444382243686252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,12288,0.04957866668701172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,12288,0.11529066827562119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,51200,0.5835706922743056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,8192,0.05952178107367622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,12288,0.14890755547417536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,10240,0.04242666562398275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,10240,0.09812711344824897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,7168,0.05158755514356825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,8192,0.03199022346072727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,10240,0.12683910793728298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,8192,0.07767733600404528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,6144,0.04013599952061971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,7168,0.029145777225494385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,7168,0.06941511233647664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,6144,0.060179558065202504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,6144,0.026027555267016094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,12288,0.08102755414115058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,7168,0.08994755480024551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,5120,0.03484977616204156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,4096,0.02897600001758999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,5120,0.02239466706911723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,10240,0.08014133241441515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,6144,0.08004888561036852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,16384,0.19533244768778482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,4096,0.019633778267436557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,4096,0.04193866584036085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,5120,0.06852977805667453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,3072,0.020779555042584736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,3584,0.018184888694021437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,4096,0.056419557995266385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,3072,0.016586666305859882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,3584,0.05115822288725111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,2560,0.017837334010336135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,8192,0.10301155514187282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,3072,0.04532177911864387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,2048,0.015298666225539314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,2560,0.015229332778188916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,2560,0.028601778878106013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,2048,0.014147554834683737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,2560,0.03877422213554382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,1536,0.011659555964999728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,2048,0.02455288834042019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,3584,0.023404444257418316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,1024,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,2048,0.033077332708570696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,1536,0.012621333201726278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,1536,0.020120888948440552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,3584,0.03748177819781833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,1024,0.015388444066047668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,1024,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,768,0.007882666256692674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,1536,0.027971555789311726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,512,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,3072,0.033161779244740806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,768,0.010393777655230628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,1024,0.022336888644430373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,768,0.013169777890046438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,256,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,512,0.011051555474599203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,512,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,128,0.0042408886882993905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,768,0.01949155496226417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,256,0.009515555368529426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,512,0.016541333662139047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,64,0.0038808890514903595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,256,0.014492443866199918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,128,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,32,0.004016889052258598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,5120,0.05082311232884725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,128,0.011349333657158745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,64,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,65536,0.25814933247036403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,32,0.008835555778609382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,51200,0.19623200098673502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,65536,0.18768889374203154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,65536,0.46989599863688153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,16384,0.07525955306159125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,51200,0.14922222826215956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,51200,0.36599821514553493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,256,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,12288,0.05494844582345751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,65536,0.5974311298794216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,16384,0.05118311113781399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,128,0.009335111412737105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,10240,0.046887109676996865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,12288,0.0403146677547031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,51200,0.4792782465616862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,16384,0.15995200475056967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,8192,0.040716442796919085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,10240,0.03566755519972907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,10240,0.0768151084582011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,12288,0.12194044060177273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,7168,0.03141866789923774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,8192,0.026211554805437725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,8192,0.06186222367816501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,10240,0.1049777799182468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,6144,0.027501333090994094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,7168,0.023749333288934495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,7168,0.05451377895143297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,5120,0.023058666123284235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,8192,0.08514489067925347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,6144,0.021393777595625982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,7168,0.07548533545600043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,4096,0.0188755558596717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,6144,0.047637333472569786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,5120,0.01899111105336083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,5120,0.040448000033696495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,16384,0.12088533242543538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,6144,0.06623911195331149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,3584,0.016976000534163583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,4096,0.016531555189026725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,4096,0.03385333220163981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,5120,0.05743733379575941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,3072,0.01464000013139513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,3584,0.015244444211324057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,12288,0.09062666363186306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,2560,0.012674666941165924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,4096,0.04821599854363335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,3584,0.030633777379989624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,3072,0.014271999398867289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,2048,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,3072,0.026644445127911035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,2560,0.01311555587583118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,3584,0.042119999726613365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,1536,0.009344889058007134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,2560,0.023230221536424425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,2048,0.011887999872366587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,3072,0.037576887342664934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,2048,0.020015999674797058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,2560,0.03247822324434916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,1536,0.010865777730941772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,2048,0.028195556667115953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,1536,0.016647110382715862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,1024,0.01000533335738712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,1024,0.012797333300113678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,1536,0.023788443870014612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,512,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,768,0.009552000297440423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,768,0.011086222198274402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,1024,0.01956266661485036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,256,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,512,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,512,0.009584888815879822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,768,0.017208000024159748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,128,0.003547555456558863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,256,0.008619555168681675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,512,0.015127110812399121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,256,0.013439111411571503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,128,0.007718222008811102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,128,0.008622222476535374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,32,0.0032213332338465583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,1024,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,128,0.010399999717871347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,64,0.0075911109646161394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,768,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,32,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,65536,0.2254284487830268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,51200,0.2007031175825331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,65536,0.18340533309512666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,65536,0.41467467943827313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,16384,0.06456000275082059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,51200,0.14636177486843532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,256,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,64,0.0031973332580592898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,12288,0.045751998821894325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,16384,0.10884088940090603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,51200,0.4817733234829373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,12288,0.03935911258061727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,16384,0.1588453319337633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,10240,0.03958666655752394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,12288,0.08232622014151679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,8192,0.02907022171550327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,10240,0.0347013341055976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,12288,0.1221724483701918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,10240,0.06882577472262912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,7168,0.026025777061780293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,8192,0.05599911345375908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,8192,0.025758221745491028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,10240,0.1043822235531277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,6144,0.022128888302379187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,7168,0.023195554812749226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,51200,0.3276195526123047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,7168,0.04931199881765577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,5120,0.019299555155966017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,6144,0.021007110675175984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,16384,0.05035555693838331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,6144,0.043014221721225314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,7168,0.07435555590523614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,65536,0.5978755421108669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,4096,0.01604711181587643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,5120,0.018665777312384713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,5120,0.03691111008326212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,6144,0.06530044476191203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,3584,0.014482667048772177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,4096,0.016308445069524977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,3072,0.012698666916953193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,4096,0.030869334936141968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,5120,0.056399113602108426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,3584,0.027650665905740526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,4096,0.04692622356944614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,2560,0.011448000040319232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,3072,0.014153778553009033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,3072,0.024380443824662104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,2048,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,3584,0.04092177748680115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,2560,0.012889777620633444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,3072,0.03637511200375027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,1536,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,2560,0.021309332715140447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,2048,0.011797332929240333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,2048,0.018446221947669983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,1024,0.006313777632183499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,8192,0.08415733443366157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,2560,0.032073779238594904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,1536,0.010713777608341642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,2048,0.02756266627046797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,1536,0.015246222416559855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,768,0.005541333307822545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,1024,0.009872888525327047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,512,0.00453333349691497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,1024,0.011757333245542316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,768,0.009546666509575313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,1536,0.0232631117105484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,256,0.003555555724435382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,1024,0.0188711103465822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,512,0.008993777963850234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,512,0.008995555341243744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,3584,0.01517777807182736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,128,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,768,0.01696622206105126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,256,0.008636444807052612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,64,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,256,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,128,0.008662222160233391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,32,0.0031662223239739737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,128,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,256,0.012739555703269111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,65536,0.1868435541788737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,64,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,128,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,32,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,51200,0.1442719962861803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,65536,0.13901778062184653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,16384,0.045144889089796275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,65536,0.33612799644470215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,51200,0.11159466372595893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,51200,0.2650737762451172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,768,0.010370666782061258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,12288,0.03353955679469638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,16384,0.039232889811197914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,65536,0.4514977667066786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,16384,0.08875377972920735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,51200,0.35730844073825413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,10240,0.026783999469545152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,12288,0.031457778480317854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,16384,0.12416622373792861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,8192,0.01993600030740102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,10240,0.05657866928312513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,10240,0.027756444282001917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,12288,0.09489155477947658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,7168,0.01720533271630605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,8192,0.019550222489568923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,8192,0.0457857780986362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,10240,0.0813973347345988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,6144,0.015059555570284525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,7168,0.01770755483044518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,8192,0.06632355848948161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,7168,0.040935112370385066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,5120,0.012768000364303589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,6144,0.03603644503487481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,6144,0.015858666764365304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,7168,0.05888533592224121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,4096,0.010976889067225985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,5120,0.014368888404634265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,6144,0.05184977915551928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,5120,0.03057510985268487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,3584,0.009931555224789513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,512,0.01482399966981676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,4096,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,4096,0.02540622154871623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,12288,0.06727644470002916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,5120,0.044574220975240074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,3072,0.008640000389681922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,3584,0.012065778175989786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,3584,0.02274399995803833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,2560,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,3072,0.011391110718250275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,4096,0.036842667394214206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,2048,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,3072,0.02049333353837331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,3584,0.03302133414480422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,2560,0.017910222212473553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,3072,0.029823111163245306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,2048,0.015497777197096082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,2048,0.010323555933104621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,2560,0.02622933354642656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,1024,0.004592000196377437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,2048,0.022853333089086745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,1536,0.012893333203262754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,768,0.0038977778620190094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,1024,0.010327999790509542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,1024,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,1536,0.019626667102177937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,512,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,1024,0.01586666703224182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,768,0.009100444614887238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,512,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,768,0.014485332700941296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,256,0.002893333426780171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,512,0.013101333545313941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,2560,0.010988444089889526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,128,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,256,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,1536,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,256,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,64,0.002513777878549364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,256,0.01109333336353302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,128,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,128,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,1536,0.00942755573325687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,65536,0.10712711016337077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,64,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,128,0.008645333349704742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,32,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,51200,0.08712710936864217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,768,0.008660444782839881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,65536,0.13784621821509466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,65536,0.29385421011182994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,16384,0.031191110610961914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,51200,0.10981866386201646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,512,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,51200,0.22978666093614367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,12288,0.02549244463443756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,16384,0.038599110311932035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,16384,0.0770346654786004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,10240,0.020190222395790946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,65536,0.438065767288208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,51200,0.34445778528849286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,12288,0.030326220724317763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,16384,0.12001066737704807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,8192,0.016467556357383728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,32,0.0027457777824666765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,10240,0.026837334036827087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,12288,0.09270399808883667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,7168,0.014793778459231058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,8192,0.019002666076024372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,8192,0.04022044605678982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,10240,0.07857333289252387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,6144,0.013171555267439948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,7168,0.01756622228357527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,7168,0.036009778579076133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,6144,0.015875554747051664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,6144,0.03177600105603536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,7168,0.056406219800313316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,5120,0.011406222151385413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,5120,0.014130666851997375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,6144,0.04994933472739326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,4096,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,5120,0.02685333291689555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,12288,0.05828977955712212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,3584,0.007213333414660559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,4096,0.012793777717484368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,4096,0.022424888279702928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,10240,0.049266666173934937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,3072,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,3584,0.012103999654452005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,3584,0.020367999871571858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,4096,0.03548355566130744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,2560,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,3072,0.018256889449225534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,3584,0.031559109687805176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,2560,0.016192000773217943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,2560,0.010980444649855295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,2048,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,3072,0.028629332780838013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,2048,0.010112000008424124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,2560,0.025342222717073228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,1536,0.004237333519591225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,2048,0.014044443766276041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,1536,0.009496889180607265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,2048,0.022032888399230108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,1024,0.003491555651028951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,1536,0.011770666473441653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,1024,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,1536,0.019088000059127808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,768,0.0031991110493739447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,1024,0.009413333402739631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,512,0.0028808888875775864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,768,0.00849955528974533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,8192,0.06392621994018555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,1024,0.01551644504070282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,768,0.008520888785521189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,256,0.0024862223201327855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,512,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,768,0.01425333321094513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,512,0.007264888948864407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,128,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,256,0.007699555820888943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,512,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,64,0.0022159999029503926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,128,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,256,0.011367999845080905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,128,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,32,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,65536,0.075254221757253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,128,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,64,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,32,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,51200,0.06252977583143446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,65536,0.13744710551367864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,65536,0.2513253423902723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,16384,0.025257777836587694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,3072,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,51200,0.11012444231245254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,51200,0.19823733965555826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,12288,0.01723377737734053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,65536,0.43349512418111164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,16384,0.03973866502443949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,16384,0.06634844674004449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,256,0.006076444354322221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,10240,0.013797333670987023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,51200,0.3404888841840956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,12288,0.031139555904600356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,12288,0.050775110721588135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,16384,0.11731466982099746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,10240,0.026171555121739704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,10240,0.04286488890647888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,7168,0.011362666885058085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,5120,0.04283555679851108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,12288,0.09182133277257283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,8192,0.01938666734430525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,6144,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,8192,0.03523644473817613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,10240,0.07677066988415189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,7168,0.01757955551147461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,5120,0.0086666668454806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,7168,0.03124444352255927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,8192,0.06239555279413859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,6144,0.01609866652223799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,6144,0.027531555957264368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,4096,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,5120,0.01443199978934394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,5120,0.023583110835817125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,6144,0.04907200071546766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,3584,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,4096,0.012776888906955719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,4096,0.019944889677895438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,5120,0.04099644554985894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,3072,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,3584,0.012146666646003723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,4096,0.03478844298256768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,3584,0.018156444032986958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,2560,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,8192,0.012936000194814471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,3072,0.011436444189813403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,3072,0.016311999824311998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,3584,0.03088444471359253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,2048,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,2560,0.010791111323568555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,2560,0.014483556151390076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,3072,0.028175999720891316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,1536,0.004028444488843282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,2048,0.012727999852763282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,2560,0.025088000628683303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,2048,0.01033688916100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,1024,0.0035084444615576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,1536,0.009658666948477427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,2048,0.0214035560687383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,7168,0.055884447362687856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,1024,0.00869688888390859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,1024,0.00869066682126787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,1536,0.01854933301607768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,512,0.002838222309947014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,1024,0.01538133290078905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,768,0.008584000170230865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,768,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,256,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,512,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,512,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,768,0.013752889302041797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,128,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,256,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,256,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,64,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,512,0.012104888757069906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,32,0.002222222172551685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,128,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,128,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,65536,0.07045066356658936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,256,0.0103493332862854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,64,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,51200,0.05642222033606636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,128,0.007991111112965478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,32,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,16384,0.022351110974947613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,12288,0.013447999954223633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,1536,0.010957332948843638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,768,0.002975111206372579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,16384,0.06629333231184217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,65536,0.25088622834947377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,12288,0.05067822337150574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,7168,0.011367999845080905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,10240,0.042809777789645724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,6144,0.009345778160625035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,5120,0.008225777910815345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,8192,0.03452533483505249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,4096,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,3584,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,7168,0.031088001198238794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,6144,0.02718488872051239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,5120,0.023433778021070693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,3072,0.005655999812814925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,2560,0.005127999931573868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,2048,0.004582222137186262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,3584,0.01812533371978336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,1536,0.003896000070704354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,3072,0.016199999385409884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,1024,0.0031600000543726813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,2560,0.01443911095460256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,768,0.002879111096262932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,1536,0.010780444575680627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,512,0.00254133323000537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,10240,0.011775111158688864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,256,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,8192,0.012471111284361945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,128,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,768,0.007879999776681265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,64,0.0021377777059872947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,512,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,51200,0.19775733682844374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,32,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,256,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,65536,0.07507999738057454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,128,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,51200,0.05734133058124118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,16384,0.020557334025700886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,32,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,12288,0.011374221907721626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,4096,0.019840000404251944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,10240,0.009855999714798397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,65536,0.25338488154941136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,16384,0.06616089079115126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,12288,0.05033777819739448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,8192,0.00869511150651508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,51200,0.19757778114742705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,2048,0.012653333445390066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,6144,0.008231999973456064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,10240,0.04270222120814853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,8192,0.03448444604873657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,7168,0.03088888857099745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,1024,0.00871022211180793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,4096,0.0087333329849773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,3584,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,6144,0.02699822187423706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,3072,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,5120,0.023585778143670823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,2560,0.006514666808976068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,4096,0.01994222237004174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,2048,0.004528000122971005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,3584,0.018210666047202218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,1536,0.0038533334930737815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,64,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,3072,0.016414221790101793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,1024,0.0032382222513357797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,768,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,2560,0.014489778214030795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,512,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,2048,0.012727110750145383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,256,0.002479111154874166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,1536,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,1024,0.008882666627566019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,768,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,7168,0.00869955536392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,64,0.0018853334089120228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,512,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,32,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,256,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,128,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,5120,0.010593777729405297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,64,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,32,0.004944889081848992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,128,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,16384,1.3260204527113173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,12288,0.9622959560818143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,12288,2.871401680840386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,16384,3.7459945678710938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,12288,3.1651734246148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,16384,4.542651706271701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,10240,0.790553781721327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,51200,4.9299439324273004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,8192,0.6573973231845432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,10240,2.960837258232964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,10240,2.5354559156629777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,8192,2.212620417277018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,8192,2.0088746812608504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,7168,0.5705617798699273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,6144,0.5033315552605523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,6144,1.618941307067871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,7168,1.7370657391018336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,6144,1.4749013053046331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,7168,2.119695027669271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,5120,0.44778580135769314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,5120,1.4542364544338648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,4096,0.44167378213670516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,4096,0.9840613471137153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,5120,1.2923217349582248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,4096,0.9676809310913086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,3584,0.3262560102674696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,3584,0.9775572882758247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,3584,0.897499508327908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,3072,0.2796391116248237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,3072,0.8701457977294922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,3072,0.7506240208943685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,2560,0.2438568803999159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,2560,0.7184106508890787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,2048,0.5163964695400661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,2048,0.20832178327772352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,2560,0.6448159747653538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,1536,0.38929955164591473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,1024,0.26563911967807347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,1536,0.18167022864023843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,2048,0.6711111068725586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,1024,0.14898755815294054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,768,0.2185786697599623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,768,0.11547822422451443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,1024,0.3030204508039686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,512,0.16087999608781603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,256,0.10599288675520156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,768,0.2865217791663276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,512,0.10537866751352946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,128,0.08020977841483222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,512,0.2484960026211209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,256,0.09156888723373413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,256,0.1903768910302056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,64,0.08447377549277411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,128,0.09032622310850356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,32,0.07695021894243029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,128,0.1730417807896932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,51200,12.269065856933594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,64,0.17568800184461805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,32,0.181823107931349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,1536,0.46008978949652773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,16384,3.1980692545572915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,51200,3.8781954447428384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,65536,5.307222154405382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,51200,19.004703097873264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,16384,1.0537911521063912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,16384,3.413428412543403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,12288,0.7655644416809082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,12288,2.2780916425916886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,51200,9.534339057074652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,12288,2.623634762234158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,10240,1.881241692437066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,10240,0.6253644625345866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,10240,2.096493403116862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,8192,0.5082969135708278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,8192,1.5374168819851344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,7168,0.4471146795484755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,7168,1.3975484636094835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,8192,1.6348710589938695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,65536,12.113381279839409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,6144,0.38805778821309406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,7168,1.33496708340115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,6144,1.1399848726060655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,5120,0.9409546322292752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,5120,0.3533955679999457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,6144,1.1973768870035808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,4096,0.7375724580552844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,4096,0.2803271081712511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,3584,0.24705065621270073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,3584,0.7136026488410102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,5120,0.9623671637641059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,4096,0.7605057822333442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,3072,0.6122737990485297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,3072,0.2450986703236898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,51200,14.086029900444878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,2560,0.5266888936360677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,2560,0.19320355521308052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,2560,0.4828622076246474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,3072,0.5852399932013618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,2048,0.17960177527533638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,2048,0.4385591083102756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,1536,0.36650223202175564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,1536,0.13142222828335232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,1024,0.23465421464708117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,768,0.17787110805511475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,1536,0.31179822815789116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,1024,0.13553067048390707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,512,0.14481422636244032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,3584,0.6671075291103787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,768,0.09436177545123631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,256,0.09675733248392741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,768,0.19814399878184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,512,0.088228444258372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,128,0.0679360032081604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,512,0.22949777709113228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,256,0.07345777750015259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,256,0.14509777228037515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,64,0.08666666348775227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,128,0.0729279981719123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,32,0.05778400103251139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,128,0.14056355423397487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,64,0.13903110557132298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,32,0.15647289488050672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,2048,0.4205626646677653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,1024,0.29079821374681264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,16384,0.9001288943820529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,51200,1.2601217693752713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,65536,1.624243524339464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,16384,0.4109768867492676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,16384,1.1171929041544597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,12288,0.9131867090861002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,51200,2.996485392252604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,12288,0.2532506783803304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,65536,19.508390638563366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,12288,0.8136062092251248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,65536,3.9724629720052085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,10240,0.2140337758594089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,10240,0.661278247833252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,8192,0.5688186751471626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,10240,0.666028446621365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,8192,0.17122666041056314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,7168,0.4651644494798448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,7168,0.19564799467722574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,6144,0.3645937707689073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,8192,0.5115884674919976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,6144,0.15668177604675293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,7168,0.5322497685750326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,6144,0.38818934228685165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,5120,0.1177333328458998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,5120,0.396013339360555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,4096,0.2791697714063856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,4096,0.09577777650621201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,5120,0.33422043588426376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,51200,4.94853761461046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,4096,0.2567111121283637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,3584,0.2697439988454183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,3072,0.20251822471618652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,2560,0.1513013309902615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,3072,0.08116266462537977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,3584,0.08938578102323745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,3584,0.2870844470130073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,2048,0.1225493351618449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,3072,0.19273332754770914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,2560,0.06706044409010145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,1536,0.1245759990480211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,2560,0.18152444892459443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,2048,0.05830578009287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,1024,0.08883377578523423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,1536,0.04745422138108147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,2048,0.14104178216722277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,1024,0.040299554665883384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,1536,0.10732355382707383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,512,0.04122666517893473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,1024,0.08553866545359294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,768,0.0373599992858039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,256,0.0302817788388994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,768,0.06612710820304023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,512,0.034168889125188194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,512,0.05636799997753567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,256,0.03058755397796631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,64,0.020040889581044514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,256,0.0506533318095737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,128,0.029490666257010564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,32,0.022633777724372015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,128,0.04816355639033847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,64,0.048635555638207324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,32,0.04928977621926201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,768,0.06027999851438734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,65536,6.375896877712673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,128,0.022249778111775715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,16384,0.6968435711330838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,65536,1.1663280063205295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,51200,0.870048893822564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,16384,0.8283528751797147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,16384,0.2590453359815809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,12288,0.19667466481526694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,12288,0.7085804409450955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,51200,2.240782207912869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,10240,0.5451911290486654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,12288,0.7031777699788412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,10240,0.20155733161502412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,65536,3.028256098429362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,8192,0.4275173346201579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,10240,0.49398310979207355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,8192,0.13252800040774876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,8192,0.4488622347513835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,7168,0.427349328994751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,6144,0.2760959996117486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,7168,0.12887467278374568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,6144,0.10394044717152913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,7168,0.3596915668911404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,5120,0.25767466757032603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,6144,0.2941466702355279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,5120,0.08941155672073364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,4096,0.20420800315009224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,3584,0.18288621637556288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,5120,0.24573421478271484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,4096,0.19598489337497285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,3072,0.14536444346110025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,3584,0.06641155481338501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,3584,0.1733928918838501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,2560,0.12941866450839573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,3072,0.05981689029269748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,3072,0.14841600259145102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,2560,0.05293155709902445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,2048,0.10098310973909165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,65536,4.485053168402778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,1536,0.07571199867460462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,2560,0.13665333059098986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,2048,0.046000887950261436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,1024,0.051557332277297974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,768,0.04347999890645345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,1536,0.042360888587103955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,1024,0.03177600105603536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,2048,0.10482488738165961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,512,0.03342666559749179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,1024,0.06435822116004096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,768,0.030626667870415583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,256,0.023157333334287006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,768,0.05151999990145365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,512,0.02755733331044515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,128,0.017613333132531907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,512,0.04338044590420193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,256,0.024359999431504145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,4096,0.08415555291705662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,64,0.016255110502243042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,256,0.03923733366860284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,128,0.02310488952530755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,32,0.016406221522225272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,51200,3.5152017805311413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,128,0.037109331952200994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,64,0.03734577695528666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,32,0.03706666827201843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,1536,0.08881333139207627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,16384,0.5469155311584473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,51200,0.757663991716173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,16384,0.21589777204725477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,65536,0.9891387091742622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,16384,0.696016894446479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,12288,0.2156764401329888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,12288,0.561174233754476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,51200,1.9098612467447917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,12288,0.5174773534138998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,10240,0.1635999944474962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,10240,0.4907502068413629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,8192,0.358603556950887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,65536,2.470993677775065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,8192,0.12461154990726048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,8192,0.3278782102796766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,7168,0.3164328998989529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,7168,0.11324177847968207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,6144,0.23417067527770996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,51200,2.907998190985786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,7168,0.307424889670478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,6144,0.08819733063379924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,5120,0.19383466243743896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,6144,0.26434577835930717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,4096,0.17352355851067436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,5120,0.07614488734139337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,3584,0.14253066645728216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,5120,0.2221591075261434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,4096,0.06318666537602742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,4096,0.1936124430762397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,3072,0.12123110559251572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,65536,3.6874597337510853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,3584,0.06155288881725735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,2560,0.0935315555996365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,10240,0.5185155338711208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,3072,0.05117866728040907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,3072,0.12747111585405138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,2048,0.08651555246777004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,1536,0.062237335575951465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,2048,0.04027022255791558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,1024,0.05054488778114319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,2048,0.08938133054309422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,1536,0.0328400002585517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,1536,0.071651558081309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,1024,0.026325333449575637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,1024,0.05432977941301134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,512,0.02849066588613722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,768,0.024284443921513025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,768,0.044360889328850635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,256,0.018772444791264005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,512,0.02217866645918952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,128,0.015860444969601102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,512,0.0359280010064443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,256,0.01976088848378923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,64,0.01332622187005149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,3584,0.14608000384436712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,256,0.03154844376775954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,32,0.013414222333166333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,128,0.019908444748984445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,2560,0.04601688848601448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,128,0.030600888861550227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,64,0.03018222252527873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,32,0.03006311257680257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,768,0.0402817792362637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,16384,0.4560231102837457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,65536,0.7771733072068957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,51200,0.6158746613396539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,16384,0.22557687759399414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,12288,0.42971022923787433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,51200,1.5751306745741103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,16384,0.5766373210483128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,65536,1.9265155792236328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,12288,0.16137688689761692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,10240,0.31416532728407115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,12288,0.5016506512959799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,8192,0.28355643484327525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,10240,0.3403280046251085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,8192,0.0938355525334676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,7168,0.21041422420077852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,8192,0.28099110391404897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,51200,2.4799395667182074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,7168,0.08422222402360703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,2560,0.11470577451917861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,6144,0.21440711286332872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,7168,0.2332515451643202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,6144,0.08327555656433105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,5120,0.17310489548577201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,4096,0.12619377507103816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,65536,3.0565821329752603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,5120,0.06747911373774211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,6144,0.2014248900943332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,3584,0.10317955414454143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,4096,0.05289333396487766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,3072,0.11021511422263251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,5120,0.16662132740020752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,4096,0.1490159961912367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,2560,0.08706488874223496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,3072,0.04217155443297493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,3584,0.1262471146053738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,2048,0.07129066520267062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,2560,0.038303110334608294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,3072,0.1042684449089898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,1536,0.05567110909356011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,1024,0.03598755598068237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,2048,0.03339733348952399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,1536,0.027035555905765955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,2048,0.07422311438454522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,768,0.02619022296534644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,1024,0.022502221994929846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,1536,0.05909866756863064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,512,0.020346666375796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,768,0.020464888877338834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,1024,0.04728177852100796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,256,0.0165964447789722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,768,0.03687466515435113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,512,0.018408000469207764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,512,0.02946044339074029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,128,0.011391110718250275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,256,0.017257778180970084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,256,0.025976889663272437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,64,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,32,0.010594666832023196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,3584,0.04942399925655789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,128,0.02471911079353756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,64,0.02506133251720005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,10240,0.11627733707427979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,2560,0.08924266364839341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,65536,0.6700586742824979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,128,0.01695644524362352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,32,0.024719999896155462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,51200,1.3173626793755424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,16384,0.38413156403435605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,65536,1.7499048444959853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,12288,0.3125964535607232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,16384,0.15737422307332358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,51200,0.6032249132792155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,16384,0.49371910095214844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,10240,0.28231289651658803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,10240,0.1350675556394789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,12288,0.37685243288675946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,51200,1.987317403157552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,8192,0.25227377149793834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,10240,0.2961920102437337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,8192,0.09972621997197469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,7168,0.1884195539686415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,65536,2.77838495042589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,6144,0.18120266331566703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,7168,0.07409333520465426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,8192,0.2635680039723714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,6144,0.06560177935494317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,7168,0.22698754734463164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,5120,0.056258665190802686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,4096,0.11698844697740342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,6144,0.18247200383080375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,3584,0.09933777650197347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,5120,0.1544426679611206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,3072,0.0830871131685045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,4096,0.12114844057295059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,3584,0.04184444414244758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,2560,0.07480177614423963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,3072,0.03786755601565043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,3072,0.09319644504123265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,2048,0.05607110924190945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,12288,0.11974667178259955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,1536,0.04337244563632541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,2048,0.028944889704386394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,2048,0.06602488623725043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,1536,0.024479111035664875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,1024,0.02853333287768894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,1536,0.05304177933269077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,768,0.0221395558781094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,1024,0.020044444335831534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,5120,0.1545191076066759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,512,0.0177839994430542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,1024,0.04057244459788004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,768,0.018255111243989732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,4096,0.04711822337574429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,256,0.01276000009642707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,512,0.016864889197879367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,128,0.011336888704035016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,512,0.026535110341178045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,256,0.01573422220018175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,3584,0.10673066642549302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,64,0.009506666825877296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,256,0.023573334018389385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,32,0.011448000040319232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,128,0.015544889701737298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,2560,0.03343200021319919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,128,0.022358222140206233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,2560,0.08006488614612155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,64,0.022685334086418152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,32,0.02237333357334137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,16384,0.34958044687906903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,768,0.03329244587156508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,51200,0.5364373524983724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,65536,0.6947626537746854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,16384,0.14936089515686035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,12288,0.29524623023139107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,16384,0.4435359901852078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,51200,1.3380080329047308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,12288,0.11156622568766277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,10240,0.2700408829583062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,12288,0.35388000806172687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,65536,1.7388106452094183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,10240,0.11437955167558457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,8192,0.2052346732881334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,10240,0.2586586740281847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,8192,0.07596088780297174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,7168,0.202801783879598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,7168,0.0722480018933614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,6144,0.1394008927875095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,8192,0.26594310336642796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,7168,0.19467911455366346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,5120,0.14430667294396296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,65536,2.3149066501193576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,6144,0.15798932976192898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,4096,0.11610489421420628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,5120,0.05232355660862393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,3584,0.10022844208611383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,4096,0.04256177941958109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,5120,0.14130932754940456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,3072,0.08449777629640366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,3584,0.040095110734303795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,3584,0.096833778752221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,2560,0.07053511010275947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,3072,0.03371555606524149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,3072,0.08127288685904609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,2048,0.05656444364123874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,2560,0.031138665146297876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,1536,0.03847555650605096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,2560,0.0705946683883667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,2048,0.026081777281231348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,1024,0.027289777994155884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,51200,1.8091742197672527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,1536,0.02204533252451155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,2048,0.05889066722657946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,768,0.020498666498396132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,1536,0.04686133397950066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,6144,0.0783404442999098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,512,0.01622222198380364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,1024,0.03589422172970242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,512,0.015511999527613321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,256,0.014876445134480795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,512,0.02347911066479153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,256,0.014533332652515836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,128,0.009308444129096137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,4096,0.11479466491275364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,256,0.02069244450993008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,64,0.008310221963458592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,32,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,128,0.01442488862408532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,128,0.01995644470055898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,64,0.01995377739270528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,32,0.01993600030740102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,1024,0.01851466629240248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,768,0.01707555519209968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,768,0.029900444878472224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,65536,0.554424868689643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,16384,0.3152799871232775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,65536,1.2519671122233074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,51200,0.4475084410773383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,16384,0.1504862176047431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,12288,0.2554444471995036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,16384,0.35944533348083496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,10240,0.20477778381771514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,12288,0.11746754911210801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,10240,0.08245511187447442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,8192,0.16595021883646646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,12288,0.31023287773132324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,10240,0.2526480091942681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,8192,0.06649421983295016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,51200,1.3858879937065973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,65536,1.8150115542941625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,8192,0.18599911530812582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,6144,0.12842755847507054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,7168,0.06303822331958346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,7168,0.15934844811757407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,5120,0.11324000358581543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,6144,0.0543813341193729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,4096,0.08169155650668673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,5120,0.04525244567129347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,6144,0.1343297825919257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,3584,0.06758755445480347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,4096,0.03885599970817566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,3072,0.0632151100370619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,3584,0.03369244601991441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,4096,0.0917288859685262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,2560,0.049386666880713574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,51200,1.1216071446736653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,3584,0.08192444509930082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,2048,0.04018311036957635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,3072,0.07282311386532254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,2560,0.027422222826215956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,1536,0.035086221165127225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,2560,0.061180445883009166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,2048,0.023330666952663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,1024,0.024673778149816725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,2048,0.05074666606055366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,1536,0.020032000210550096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,768,0.020057777563730877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,1536,0.0410755541589525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,1024,0.017504000001483493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,512,0.014934221903483072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,1024,0.03077510992685954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,768,0.016145777371194627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,256,0.01055555542310079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,768,0.026152888933817547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,128,0.008932444784376357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,512,0.020750222934616935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,256,0.018218666315078735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,64,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,128,0.013448889056841532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,5120,0.11799377865261501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,128,0.01752088963985443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,32,0.007827555967701806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,3072,0.03037244412634108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,64,0.017536888519922893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,32,0.01750666730933719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,7168,0.15997332996792263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,65536,0.45207643508911133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,512,0.01442577772670322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,256,0.01347555551264021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,65536,0.9217661751641167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,51200,0.6896177927652994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,16384,0.2242666615380181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,16384,0.10382933086819118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,12288,0.20576266447703043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,51200,0.3505404525332981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,16384,0.29044800334506565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,12288,0.07998844650056627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,10240,0.1787386602825589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,12288,0.21800355116526285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,10240,0.06799199846055773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,8192,0.14787467320760092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,65536,1.5990524291992188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,10240,0.22453512085808647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,7168,0.131296886338128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,8192,0.05480622251828512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,51200,1.2150782479180229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,6144,0.10700800021489461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,7168,0.05713955561319987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,8192,0.15405155552758112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,6144,0.04253066579500834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,7168,0.1298497782813178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,6144,0.11155200004577637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,3584,0.05592266718546549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,4096,0.030701332622104224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,5120,0.1017964416080051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,4096,0.07662844657897949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,3072,0.0485173331366645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,3584,0.02837688724199931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,2560,0.04148177636994256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,3584,0.06728977627224393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,3072,0.02551377813021342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,2048,0.034584889809290566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,2560,0.02280355493227641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,3072,0.060828447341918945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,1536,0.02602133320437537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,2560,0.051926222112443715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,2048,0.020083554916911654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,1024,0.018811555372344125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,2048,0.0437520013915168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,1536,0.017458667357762653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,768,0.014808000789748298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,1536,0.03469333383772109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,512,0.011352889239788055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,1024,0.025846223036448162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,768,0.013996443814701505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,256,0.008435555630260045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,5120,0.08764444457160102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,768,0.022294221652878657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,512,0.012852444416946836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,128,0.006909333169460297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,512,0.018207111292415194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,256,0.012073777616024017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,5120,0.03770044445991516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,64,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,32,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,256,0.015511999527613321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,128,0.012131555212868584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,128,0.01517599986659156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,64,0.01515111161602868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,32,0.014816888504558139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,65536,0.40889689657423234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,51200,0.32827554808722603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,1024,0.015145777000321282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,65536,0.9238932927449545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,51200,0.7596382035149468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,16384,0.24829777081807455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,16384,0.09631999995973374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,4096,0.06877155436409844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,12288,0.19701511330074736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,12288,0.08429333236482407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,10240,0.16882667276594374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,16384,0.2876524395412869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,10240,0.07545066542095609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,8192,0.13886133829752603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,65536,1.2797760433620877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,8192,0.05202666587299771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,10240,0.16378488805558947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,6144,0.10507200161616008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,7168,0.04809866680039299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,5120,0.08311733272340563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,7168,0.11923466788397895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,6144,0.03975288735495673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,6144,0.10131822029749553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,4096,0.06690222024917603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,5120,0.03440711233350966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,3584,0.05736177497439914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,4096,0.029333333174387615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,4096,0.06906755765279134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,3584,0.02564977771706051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,3584,0.061055110560523145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,3072,0.049238221512900464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,51200,0.998527103000217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,3072,0.023056000471115112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,12288,0.26269955105251735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,2560,0.04093866546948751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,7168,0.11760177877214219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,2048,0.028421332438786823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,8192,0.13282844755384657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,1536,0.023698666029506262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,2560,0.04675288995107015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,2048,0.039001779423819646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,1024,0.01573066744539473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,1536,0.016150222884284127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,1536,0.03177155719863044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,768,0.012881778180599213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,1024,0.013809777796268463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,5120,0.08936799897087945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,512,0.010715555813577441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,1024,0.02347555591000451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,768,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,256,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,768,0.020233778489960563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,512,0.01221244451072481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,128,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,512,0.01640888883007897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,256,0.011751111182901593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,64,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,256,0.014431110686726041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,128,0.011399110986126794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,3072,0.053710222244262695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,32,0.006572444405820634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,128,0.013784888717863293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,2560,0.02092711130777995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,64,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,2048,0.01835466590192583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,32,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,16384,0.191703995068868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,65536,0.36116798718770343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,51200,0.2799253463745117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,65536,0.7347928682963053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,16384,0.08672000302208795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,12288,0.17034844557444254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,16384,0.22938844892713758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,12288,0.07311289178000556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,10240,0.1317288875579834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,65536,1.0338470670912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,51200,0.8203582233852811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,8192,0.11077155669530232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,10240,0.05790044201744927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,7168,0.0984684427579244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,8192,0.05068888929155138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,6144,0.07680888970692952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,8192,0.12534933620029026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,7168,0.04100799891683791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,5120,0.06695199675030179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,7168,0.1075893309381273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,4096,0.05304000112745497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,6144,0.08989777829911973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,5120,0.07670755518807305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,4096,0.026272000537978277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,51200,0.6850302484300402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,3584,0.023875556058353845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,3584,0.05351022217008802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,3072,0.03711288836267259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,12288,0.18610221809811064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,2560,0.032840887705485024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,3072,0.02159822152720557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,10240,0.14515554904937744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,3072,0.047305779324637524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,2048,0.025780444343884785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,2560,0.01926399932967292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,1536,0.020989333589871723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,6144,0.03894311189651489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,2048,0.01737777723206414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,2560,0.041124443213144936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,1024,0.016169778174824186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,2048,0.03446755475468106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,5120,0.03179644544919332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,768,0.012793777717484368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,1024,0.013096888860066732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,1536,0.027912888262006972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,3584,0.04423200090726217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,512,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,1024,0.02119111186928219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,768,0.012616888516479068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,256,0.0073520003093613525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,4096,0.06115555763244629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,128,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,512,0.011709333293967776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,256,0.011071999867757162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,64,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,512,0.0148026661740409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,128,0.011050666371981302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,256,0.012898666991127862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,128,0.012484444512261285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,64,0.012392000191741519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,32,0.01275288893116845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,65536,0.5256231096055772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,65536,0.3058497905731201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,51200,0.4980417887369792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,1536,0.015103111664454142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,16384,0.16036178006066218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,51200,0.2418924437628852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,768,0.0181040002240075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,65536,0.8828204472859701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,12288,0.12304266293843587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,51200,0.6739919980367025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,16384,0.07765333520041572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,16384,0.20011733637915718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,32,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,10240,0.10632266600926717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,12288,0.06069066789415148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,8192,0.0800515545739068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,10240,0.05420088768005371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,7168,0.07628977961010404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,12288,0.15123911698659262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,8192,0.04267911116282145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,6144,0.05915288792716133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,10240,0.13201510906219482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,7168,0.036638223462634616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,8192,0.10281955533557469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,6144,0.03259377678235372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,7168,0.09135911199781631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,4096,0.037976887491014265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,6144,0.07944355408350627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,5120,0.028019554085201685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,3584,0.03680000040266249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,5120,0.06648977597554524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,3584,0.021719111336602107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,4096,0.055085334512922496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,2560,0.028363555669784546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,3584,0.04849777619043986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,3072,0.01941866676012675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,3072,0.042845331960254245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,2048,0.023823999696307715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,2560,0.017575111654069688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,1536,0.01941244469748603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,2048,0.015849777393870883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,2048,0.03141244583659702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,1024,0.014364444547229342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,1536,0.025225778420766194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,5120,0.04570133487383524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,1024,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,768,0.011809777882364062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,1024,0.019073777728610568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,512,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,768,0.011683555940786997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,4096,0.023963555693626404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,256,0.0063457778758472866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,512,0.011006222003036074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,512,0.013460444079505073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,256,0.01036977767944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,256,0.011776888536082374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,64,0.005040888984998067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,128,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,32,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,128,0.011386666860845355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,2560,0.03689600030581156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,64,0.011100444528791638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,1536,0.0144213338692983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,32,0.011362666885058085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,65536,0.5061982472737631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,65536,0.276422209209866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,51200,0.39484087626139325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,3072,0.03394044438997904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,768,0.016182222300105624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,16384,0.13420622878604466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,12288,0.09464799695544773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,51200,0.22966400782267252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,65536,0.7166328959994845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,16384,0.07107555866241455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,51200,0.5543440183003744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,10240,0.08244888650046454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,16384,0.18878400325775146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,8192,0.07343555821312799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,12288,0.1295902199215359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,7168,0.06788177622689141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,8192,0.03886222177081638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,8192,0.09101066986719768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,6144,0.05652444230185615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,7168,0.03310399916436937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,5120,0.04636711213323805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,7168,0.07777777645323011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,6144,0.03016977839999729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,4096,0.034182220697402954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,6144,0.06840888659159343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,5120,0.05650933583577474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,3584,0.030798223283555772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,4096,0.02183733383814494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,4096,0.0469777782758077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,3072,0.028014222780863445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,3584,0.020017777880032856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,3584,0.04184800055291918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,2560,0.02331377731429206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,3072,0.017204445269372728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,12288,0.05551288856400383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,3072,0.03700622253947788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,10240,0.048729777336120605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,2560,0.015790222419632804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,10240,0.11241688993242051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,1536,0.014589332871966891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,2560,0.032051556640201144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,2048,0.01519377695189582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,2048,0.026800000005298193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,1024,0.010990222295125326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,1536,0.01276622215906779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,768,0.00868711123863856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,1536,0.021928000781271193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,1024,0.011328889264000786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,5120,0.026200888885392085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,512,0.00721866637468338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,1024,0.016698666744761996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,256,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,768,0.010686222049925061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,768,0.014492443866199918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,512,0.010008888940016428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,128,0.0046897778908411665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,256,0.009815110928482479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,64,0.0042151109211974675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,128,0.005535999933878581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,128,0.009336888790130615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,256,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,32,0.004517333375083076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,128,0.01015822258260515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,2048,0.01906399925549825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,64,0.00981866651111179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,32,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,65536,0.22594579060872397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,51200,0.33774577246771914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,16384,0.10787377754847209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,51200,0.18017244338989258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,12288,0.08078666528065999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,65536,0.5540497567918565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,16384,0.059711111916436084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,51200,0.43944088617960614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,10240,0.06905421945783827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,12288,0.047361777888404004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,16384,0.15311733881632486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,8192,0.05554844604598152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,12288,0.10645600159962972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,10240,0.041152887874179415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,7168,0.04639288783073425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,10240,0.0887884431415134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,8192,0.031361778577168785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,6144,0.03981599873966641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,8192,0.07494133048587374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,5120,0.033619556162092425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,7168,0.028491556644439697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,6144,0.025166221790843542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,7168,0.06330221891403198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,5120,0.022026666336589392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,6144,0.055276443560918175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,3584,0.02388266722361247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,5120,0.04764799939261543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,65536,0.42299289173550075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,3072,0.020454222957293194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,4096,0.03859377900759379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,2560,0.01760444376203749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,3584,0.034685333569844566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,3072,0.031007111072540283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,2048,0.014894222219785055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,2560,0.014832889040311178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,2560,0.02659822172588772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,1536,0.011744889120260874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,2048,0.01352533366945055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,512,0.012079111403889127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,2048,0.022572444544898138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,1024,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,1536,0.012201777762836881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,1536,0.018574222922325134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,768,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,1024,0.01089422239197625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,512,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,1024,0.014462222655614218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,768,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,256,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,768,0.012479999827014076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,512,0.009658666948477427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,4096,0.018978666928079393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,512,0.01053155544731352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,256,0.009189333352777693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,256,0.00908799966176351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,64,0.004223111189073986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,32,0.00434666665063964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,3072,0.016189333465364244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,128,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,64,0.008600888980759515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,32,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,65536,0.29979467391967773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,51200,0.20451911290486655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,65536,0.17938044336107042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,16384,0.0747599999109904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,65536,0.44183733728196883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,4096,0.026594666971100703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,51200,0.14408977826436362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,51200,0.35023021697998047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,12288,0.05310755636956957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,10240,0.04125510983996921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,16384,0.049004445473353066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,128,0.004754666652944353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,12288,0.03907555672857497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,16384,0.11540267202589248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,3584,0.017539555827776592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,8192,0.03684444559945001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,7168,0.03239466746648153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,12288,0.08680799934599136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,128,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,10240,0.03425777620739407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,6144,0.027338667048348322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,8192,0.025582222474945918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,10240,0.0726755526330736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,5120,0.02293066680431366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,7168,0.023222222924232483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,8192,0.05929866764280531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,4096,0.01886311173439026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,7168,0.05252088771926033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,5120,0.018386666973431904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,3584,0.01686044368478987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,6144,0.0457857780986362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,3072,0.01477688882086012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,4096,0.016190222567982145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,5120,0.038859556118647255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,2560,0.012748444245921241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,4096,0.032411555449167885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,3584,0.01501511037349701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,2048,0.010791999598344168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,3072,0.013816888961527081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,3584,0.029137776957617864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,2560,0.012719111310111152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,3072,0.025979555315441553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,2560,0.02253955602645874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,2048,0.011721777419249216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,1024,0.0070053330726093715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,2048,0.01943555474281311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,1536,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,768,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,1536,0.016151999433835346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,1024,0.009683555199040307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,512,0.005199999858935674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,1024,0.012401777837011548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,768,0.009542221824328104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,256,0.004566222015354368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,768,0.01088711122671763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,512,0.008958222137557136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,128,0.0038444445365005066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,512,0.009319999979601966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,64,0.003519999898142285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,256,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,32,0.00350222239891688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,6144,0.021205332544114854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,128,0.007612444460391998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,64,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,65536,0.2071919971042209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,32,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,51200,0.1711475584242079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,65536,0.1772435506184896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,1536,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,16384,0.05922400289111667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,51200,0.14219644334581164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,65536,0.36830933888753253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,12288,0.04645422101020813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,10240,0.03563111027081808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,16384,0.048568000396092735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,51200,0.2945511076185438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,12288,0.03862222366862827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,8192,0.027540443672074214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,16384,0.09623644749323527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,12288,0.07230844762590197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,10240,0.03345866666899787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,7168,0.02561155623859829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,6144,0.02237333357334137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,8192,0.025476444098684523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,10240,0.061365332868364125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,5120,0.01942222151491377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,7168,0.02281422250800663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,8192,0.0498542222711775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,4096,0.016353777713245816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,256,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,7168,0.044250667095184326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,128,0.008471999731328752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,5120,0.018256000346607633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,3584,0.01423466702302297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,6144,0.038775109582477145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,3072,0.012308444413873883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,4096,0.01590666671593984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,5120,0.033313777711656355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,2560,0.011040888726711273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,3584,0.014829332629839579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,4096,0.027590221828884546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,3072,0.013802666631009845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,2048,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,3584,0.024725332856178284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,1536,0.007822222179836696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,3072,0.02198133369286855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,2560,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,1024,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,2048,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,2560,0.01924888955222236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,1536,0.010670222342014313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,768,0.005518222020732031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,2048,0.01660355594423082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,1024,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,1536,0.013814222481515674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,768,0.009380444056457942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,256,0.0042204442951414315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,768,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,512,0.008295999632941352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,256,0.008377778033415476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,128,0.0035182221068276297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,256,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,64,0.0028844444702068963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,6144,0.02066044509410858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,128,0.008391111261314815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,32,0.003154666680428717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,128,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,64,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,65536,0.1933679978052775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,32,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,51200,0.14213421609666613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,65536,0.1327457825342814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,65536,0.31863289409213597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,512,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,16384,0.0488702224360572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,1024,0.010999999940395355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,51200,0.25189420912000865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,512,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,16384,0.037478221787346735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,10240,0.028759111960728962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,12288,0.028925332758161757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,16384,0.08485155635409886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,8192,0.022218666142887537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,12288,0.06387022468778822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,10240,0.0258240004380544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,7168,0.0203804439968533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,10240,0.05387644304169548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,8192,0.01865955524974399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,6144,0.01738577749994066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,8192,0.04359822140799629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,7168,0.017248888810475666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,5120,0.01532444523440467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,7168,0.039192001024881996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,6144,0.015778667396969266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,4096,0.012551111479600271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,3584,0.011558222273985544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,6144,0.03436622354719374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,4096,0.012718222207493253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,3072,0.010192888478438059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,3584,0.011750222080283694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,4096,0.024238222175174292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,2560,0.008736000292830998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,3584,0.02197866638501485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,3072,0.011095999843544431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,12288,0.035990221632851496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,2048,0.007721777591440413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,3072,0.019906666543748643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,2560,0.010674667027261523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,2560,0.017426666286256578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,2048,0.009992889232105678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,1024,0.00519466648499171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,2048,0.014871999621391296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,1536,0.009393778112199571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,768,0.00452622233165635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,1536,0.01272266689274046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,1024,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,512,0.0038773334688610504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,768,0.008610666626029545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,768,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,512,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,5120,0.014157333307796054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,512,0.007693332930405934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,128,0.002847111059559716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,256,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,256,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,64,0.002536888958679305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,128,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,128,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,51200,0.11254844400617812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,32,0.0027928888383838865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,1536,0.006451555424266391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,65536,0.1166435612572564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,64,0.006218666831652324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,51200,0.09257511297861735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,65536,0.13202222188313803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,65536,0.2764035595787896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,51200,0.10572622219721477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,51200,0.21737155649397108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,12288,0.02318488889270359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,256,0.003160888950030009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,16384,0.03714133302370707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,16384,0.07202044460508558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,10240,0.019532443748580087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,5120,0.02941511074701945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,12288,0.028889778587553237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,8192,0.01612622208065457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,10240,0.02517866591612498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,12288,0.05505333344141642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,7168,0.014338667194048563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,10240,0.04655999938646952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,8192,0.03811111052831014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,7168,0.017160000072585214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,6144,0.012839111189047495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,7168,0.03404444456100464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,5120,0.011376000112957425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,32,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,6144,0.015781333049138386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,4096,0.009443555441167619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,6144,0.030072000291612413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,5120,0.01425066590309143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,5120,0.02553066611289978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,1024,0.00998222248421775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,3584,0.008652444514963362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,4096,0.012517333030700684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,3072,0.007639110916190677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,4096,0.021427555216683283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,2560,0.0070222218831380205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,3584,0.011701333026091257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,3072,0.011130666567219628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,2048,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,3072,0.017664000391960144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,2560,0.010478222535716163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,1536,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,2560,0.01552177800072564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,2048,0.010059555371602377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,1024,0.00453955555955569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,2048,0.01348622226052814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,1536,0.009379555781682333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,768,0.0038355555799272326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,1536,0.011376889215575324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,1024,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,512,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,768,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,1024,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,768,0.008310221963458592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,256,0.002833777831660377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,512,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,512,0.007204444044166141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,256,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,256,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,64,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,128,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,32,0.002483555633160803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,3584,0.019497777024904888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,65536,0.07281155718697442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,64,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,32,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,51200,0.058993776639302574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,8192,0.018586667047606576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,16384,0.024382222029897902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,65536,0.13217422697279188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,51200,0.10591910945044623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,65536,0.23532888624403211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,12288,0.0158906661801868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,16384,0.03869333200984531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,16384,0.06252711349063449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,10240,0.01405422223938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,16384,0.03206133180194431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,128,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,8192,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,10240,0.02644799980852339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,12288,0.04769688844680786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,7168,0.011571555501884885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,10240,0.04004888733228048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,8192,0.018920888503392536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,6144,0.010112000008424124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,5120,0.00905866672595342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,8192,0.032976888948016696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,7168,0.029422223567962646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,6144,0.015831111205948722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,4096,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,5120,0.014320888453059726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,3584,0.00703111125363244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,5120,0.02235377828280131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,4096,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,3072,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,4096,0.018956444329685636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,3584,0.012439111040698158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,2560,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,3072,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,3584,0.01724799970785777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,51200,0.1858888864517212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,2048,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,3072,0.015784889459609985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,1536,0.004517333375083076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,2560,0.01073333372672399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,2048,0.010025777750545079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,2560,0.013844444519943662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,1024,0.0035333335399627686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,1536,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,2048,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,768,0.0031662223239739737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,1536,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,1024,0.008624888956546783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,512,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,768,0.008374222450786168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,1024,0.008402666283978356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,7168,0.018264888061417472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,512,0.00794044468137953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,512,0.0068195557428730865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,256,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,6144,0.025892444782786902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,128,0.0022151112142536375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,64,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,256,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,32,0.0022355556074115965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,128,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,128,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,65536,0.07395733065075345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,64,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,51200,0.05284888876809014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,32,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,16384,0.01867377758026123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,12288,0.01236088905069563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,12288,0.030004445049497817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,10240,0.011037333144081963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,65536,0.2366960048675537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,16384,0.06251733170615302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,51200,0.18543733490837944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,8192,0.009719111025333405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,7168,0.009201777478059133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,12288,0.04748622245258755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,6144,0.009324444664849175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,5120,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,10240,0.04007733199331496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,768,0.00795022232664956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,8192,0.032609777318106756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,4096,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,7168,0.029105779197480943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,3584,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,6144,0.025747555825445387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,2560,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,5120,0.022286223040686712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,2048,0.004545777622196409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,4096,0.018927999668651156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,1536,0.0039057777159743835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,3584,0.01720977822939555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,1024,0.0032231110251612137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,3072,0.015639111399650574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,768,0.002874666617976295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,512,0.002826666666401757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,2560,0.013825777504179211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,256,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,2048,0.012113778127564324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,1536,0.010431110858917236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,128,0.0021333334346612296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,64,0.0021119999388853707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,512,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,32,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,256,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,128,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,65536,0.06980444325341119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,64,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,51200,0.055352889829211764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,32,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,16384,0.017442666822009616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,256,0.00254133323000537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,12288,0.010732444624106089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,10240,0.009681777821646797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,3072,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,65536,0.23658400111728242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,51200,0.18551466200086805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,8192,0.008635555704434713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,16384,0.06199200285805596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,12288,0.047469334469901196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,6144,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,10240,0.03997333182228936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,5120,0.011591999895042844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,4096,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,8192,0.03263911273744371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,7168,0.028950221008724634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,6144,0.0259244441986084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,3584,0.008799110849698385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,3072,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,1024,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,2560,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,768,0.007637333538797166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,4096,0.018927110566033255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,2048,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,1536,0.003863111138343811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,1024,0.0031715554909573663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,3584,0.01718755563100179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,3072,0.015525332755512662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,2560,0.013783999615245394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,2048,0.012135110795497894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,512,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,1536,0.010398222340477837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,256,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,1024,0.00847466703918245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,768,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,64,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,512,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,256,0.005668444352017508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,32,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,128,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,7168,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,64,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,32,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,5120,0.02232711182700263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,768,0.002871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,128,0.0021084443562560612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,16384,3.675279829237196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,16384,1.181293381585015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,16384,4.279487186008029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,51200,4.596922556559245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,16384,3.783037397596571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,12288,2.918778737386068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,12288,0.9148506588406033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,12288,3.3153076171875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,10240,2.5362364451090493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,10240,0.7710391150580512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,12288,2.8079590267605252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,10240,2.5237467024061417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,8192,0.5968009101019965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,10240,2.33915032280816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,8192,2.0006817711724176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,8192,2.0479244656032987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,51200,11.871853298611112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,8192,1.9266088273790147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,7168,0.5457937982347276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,7168,1.715669314066569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,7168,1.7414408789740667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,6144,0.468556457095676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,6144,1.6697582668728297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,7168,1.6200977961222331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,6144,1.422120836046007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,5120,1.3900897767808702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,5120,0.4004053274790446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,6144,1.3848240110609267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,4096,0.8631359736124674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,5120,1.2640239927503798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,4096,0.33485065566168887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,3584,0.7876826922098795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,4096,1.0039964252048068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,5120,1.1140577528211806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,3584,0.8439075681898328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,4096,0.9021004570855035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,3584,0.3026817904578315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,3072,0.6955777804056803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,3072,0.27081865734524196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,3584,0.7970799869961209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,3072,0.7408089107937283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,2560,0.5843360159132216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,3072,0.6767191357082791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,2560,0.23564889695909289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,2048,0.47396442625257706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,2560,0.6999937693277994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,2560,0.5746542082892524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,2048,0.19579733742607963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,2048,0.5535590913560655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,1536,0.39185510741339785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,1536,0.15818577342563206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,1024,0.26289600796169704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,2048,0.4673617680867513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,1536,0.39252977901034886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,768,0.20428266790178087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,1024,0.14606489075554743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,1024,0.3044755458831787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,768,0.23385243945651582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,1536,0.3728293312920465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,51200,13.272151523166231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,512,0.14933510621388754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,1024,0.27236445744832355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,768,0.1309600008858575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,256,0.11260888311598037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,512,0.12311377790239121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,768,0.23508800400627983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,51200,18.964012993706596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,256,0.09813688860999213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,256,0.17932444148593477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,128,0.11529600620269775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,512,0.19994133048587373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,512,0.18858044677310518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,32,0.07122133175532024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,256,0.14558045069376627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,128,0.09029066562652588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,128,0.17113333278232148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,128,0.11819110976325141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,32,0.17489066388871935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,64,0.1774515575832791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,64,0.06975199778874715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,51200,3.5786416795518665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,65536,4.721731397840712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,16384,3.6322284274631076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,51200,9.501895480685764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,16384,3.5803646511501737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,16384,0.9035689036051432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,65536,12.036156548394096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,12288,2.614114761352539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,12288,0.6870089107089572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,16384,2.9996560414632163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,12288,2.520047081841363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,51200,14.510272555881075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,51200,10.57384999593099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,12288,2.133278316921658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,10240,0.602199978298611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,10240,2.321271048651801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,10240,2.113759994506836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,8192,1.533498658074273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,8192,0.5811253123813206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,10240,1.8381031884087458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,8192,1.5512666702270508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,7168,1.4232070710923936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,65536,17.88254123263889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,7168,0.4286862214406331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,8192,1.4432053036159938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,65536,13.0164794921875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,6144,0.37319376733567977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,7168,1.3804551230536566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,6144,1.125788476732042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,7168,1.2729715771145291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,6144,1.1285200119018555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,5120,0.9099440044826932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,5120,0.3142186535729302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,6144,1.0628044340345595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,4096,0.6474391089545356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,4096,0.2668986585405138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,5120,0.9498693678114148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,5120,0.8557030889723035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,4096,0.7738169034322103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,3584,0.687387572394477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,4096,0.6834159957038032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,3584,0.24149600664774576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,3072,0.5245928764343262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,3072,0.2547653251224094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,3584,0.6798382335238986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,2560,0.4760755432976617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,3584,0.6054782337612575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,3072,0.5793457561069065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,2048,0.3313804467519124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,2560,0.22715200318230522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,2560,0.5088595814175075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,3072,0.5338142183091905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,2048,0.1554515494240655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,1536,0.3085671001010471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,1024,0.19440710544586182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,1536,0.12627555264367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,2048,0.48718754450480145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,1536,0.32510222329033744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,2048,0.36980178621080184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,768,0.1488915549384223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,1024,0.10053866439395481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,1536,0.2950870990753174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,512,0.11401244004567464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,1024,0.35340176688300234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,768,0.18546666039360893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,1024,0.21760177612304688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,256,0.07795644468731351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,512,0.08971199724409316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,768,0.08919288714726765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,768,0.1777813302146064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,128,0.06288088692559136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,512,0.158879107899136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,256,0.07166488965352376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,256,0.14299910598331028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,512,0.1435182227028741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,64,0.058394663863711886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,32,0.05748266643948025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,256,0.11337778303358291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,128,0.13434933291541204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,2560,0.44962490929497617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,128,0.08685777584711711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,64,0.13627110587226018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,32,0.1379502216974894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,128,0.06846488846672906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,65536,1.4703324635823567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,51200,1.114715576171875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,16384,1.328378677368164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,51200,3.2767431471082897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,65536,4.058180491129558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,16384,0.30910666783650714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,16384,1.2428009245130751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,12288,0.9076515833536783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,51200,3.2209538353814016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,12288,0.7869413163926867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,16384,0.9406568739149305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,51200,5.2383575439453125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,12288,0.2648373444875081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,12288,0.6642533408270942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,10240,0.20736889044443765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,10240,0.6719537840949165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,10240,0.6366711192660862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,65536,4.216102176242404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,8192,0.5649689038594564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,8192,0.1942666636572944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,10240,0.5654222170511881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,7168,0.3937226666344537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,8192,0.5155839920043945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,65536,6.6166593763563375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,7168,0.18096889389885795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,8192,0.4408764574262831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,6144,0.3567057715521918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,6144,0.12647644678751627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,5120,0.30757154358757866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,7168,0.58225064807468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,5120,0.10898755656348334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,6144,0.41738221380445695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,7168,0.5385226673550075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,4096,0.24285689989725748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,5120,0.3263733386993408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,6144,0.3420791096157498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,4096,0.09151199791166519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,4096,0.2443324459923638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,5120,0.28607021437750924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,3584,0.08214844597710504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,3072,0.16142666339874268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,3584,0.22211289405822754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,4096,0.23088532023959688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,2560,0.15219377146826849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,3072,0.19215022193060982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,3584,0.20746133062574598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,2048,0.13851466443803576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,2560,0.06462044186062284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,2560,0.16188177797529432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,3072,0.17946488327450225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,2048,0.05521688858668009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,1536,0.09941511021720038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,2560,0.15421777301364475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,2048,0.13391466935475668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,1024,0.05894133117463854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,1536,0.04547199938032362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,3584,0.24941778182983398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,2048,0.1298497782813178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,1536,0.10572444730334812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,768,0.05064177844259474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,1024,0.03801066676775614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,1024,0.08125333653555976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,512,0.03556888964441087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,768,0.03511111272705926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,3072,0.07314133644104004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,1024,0.08166844314999051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,256,0.028397334946526423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,768,0.06840622425079346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,512,0.03216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,512,0.055530667304992676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,128,0.02072533302836948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,256,0.02889510989189148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,256,0.049902220567067467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,512,0.05560533205668131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,64,0.0188426673412323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,32,0.018897778458065458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,256,0.04424533247947693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,128,0.04714666803677877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,128,0.028218666712443035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,64,0.047819554805755615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,128,0.03444000085194906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,32,0.04850133260091146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,1536,0.10351288980907863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,768,0.06506222486495972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,65536,1.0445653067694771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,51200,0.8238417837354871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,51200,2.1885723537868924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,16384,0.9439155790540906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,65536,2.805723614162869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,16384,0.9260390599568685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,16384,0.26154398918151855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,12288,0.5482915772332085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,51200,3.382633845011393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,12288,0.5870684517754449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,12288,0.19440889358520508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,51200,2.5351325141059027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,10240,0.4310622215270996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,12288,0.5134506755405003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,65536,4.169840070936415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,10240,0.16445688406626383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,10240,0.49611642625596786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,65536,3.2436996036105685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,8192,0.37576977411905926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,8192,0.12512711683909097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,10240,0.4275875621371799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,7168,0.33281156751844615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,8192,0.40712089008755153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,7168,0.1128586663140191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,16384,0.6974035369025336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,6144,0.2687031163109673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,8192,0.34563644727071124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,7168,0.3609457810719808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,6144,0.09891200065612793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,5120,0.2365626758999295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,6144,0.3322319984436035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,7168,0.30776000022888184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,4096,0.19392711586422393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,6144,0.2626444498697917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,3584,0.1565271086162991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,4096,0.07307910919189453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,4096,0.19700088765886095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,3584,0.06421066655053033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,3072,0.129838228225708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,3584,0.1688222222858005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,4096,0.1827911138534546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,2560,0.11533421940273708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,3584,0.16450133588578966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,3072,0.05765599674648709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,3072,0.1463137732611762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,2560,0.050424890385733716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,2048,0.0806808869043986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,3072,0.14139200581444636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,2560,0.1258702278137207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,1536,0.06718755430645414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,5120,0.0844871136877272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,2048,0.043473776843812734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,2560,0.12120532989501953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,5120,0.2554275459713406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,2048,0.10322133037779067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,1024,0.05233155687650045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,1536,0.036313778824276395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,1536,0.08261689212587145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,2048,0.10411200258466934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,768,0.03734133309788174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,1024,0.06335110796822442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,1024,0.030619555049472388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,1536,0.08358133501476711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,512,0.02941155433654785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,1024,0.06570577621459961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,768,0.050683554675843984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,256,0.02011022302839491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,512,0.025898666845427618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,512,0.04308533337381151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,768,0.05581600136227078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,128,0.018960000740157235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,256,0.023410666320059035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,128,0.036313778824276395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,256,0.03630133469899496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,64,0.016160888804329764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,128,0.023002665903833177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,32,0.015990222493807476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,128,0.02790133489502801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,64,0.036706666151682533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,32,0.036806222465303205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,768,0.028559111886554297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,256,0.038520889149771795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,5120,0.22574133343166777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,65536,0.9246435165405273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,512,0.04580444428655836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,65536,2.543627632988824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,51200,0.770726203918457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,16384,0.6678106519911023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,51200,2.1767893897162542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,16384,0.20607022444407144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,12288,0.5146035618252224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,16384,0.6734515296088325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,51200,2.817721684773763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,12288,0.15920444329579672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,65536,3.6453696356879344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,12288,0.4943191210428874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,65536,2.6675777435302734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,10240,0.37149688932630753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,51200,2.002265718248155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,10240,0.1332657734553019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,12288,0.4260266621907552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,10240,0.3984995683034261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,8192,0.33792800373501247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,8192,0.10813333590825398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,7168,0.27863910463121205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,10240,0.36651910675896543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,7168,0.09667022360695733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,7168,0.28077954716152614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,6144,0.2421893278757731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,8192,0.3707919915517171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,16384,0.5999546580844456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,7168,0.2952115535736084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,5120,0.20074933105044893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,4096,0.16322755813598633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,5120,0.07407377825842963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,6144,0.21913511223263213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,5120,0.2153093285030789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,4096,0.060588445928361684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,3584,0.1192515558666653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,5120,0.18796800242529976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,4096,0.1953484482235379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,8192,0.3412951098548041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,3072,0.10960088835822211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,4096,0.15503733687930638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,3584,0.1408888896306356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,3072,0.04837333493762546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,2560,0.09318133195241292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,3584,0.13834933439890543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,3072,0.12342577510409886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,6144,0.24704355663723418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,2048,0.07765599754121569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,3072,0.12214044729868571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,2560,0.043634666336907275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,1536,0.06516177786721124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,2048,0.03785422113206651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,2048,0.08857155508465236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,1536,0.0314702226055993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,2048,0.08912089135911729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,1536,0.0699582232369317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,1024,0.04487911197874281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,3584,0.05499022205670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,1536,0.07284888956281874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,1024,0.02517333295610216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,768,0.033702221181657575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,1024,0.053447110785378345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,512,0.02423733307255639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,768,0.023413333627912734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,1024,0.05670577949947781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,768,0.043255110581715904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,256,0.01792888840039571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,512,0.02130399975511763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,512,0.03542133172353109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,768,0.04791288905673557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,2560,0.10666310787200928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,128,0.014125333891974555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,256,0.019137778215938143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,512,0.0395084453953637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,2560,0.1040266686015659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,256,0.030973331795798406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,64,0.013636444177892474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,32,0.01622577839427524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,128,0.018933332628673978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,256,0.031927112076017596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,64,0.029848890172110662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,32,0.029503110382292006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,6144,0.08503199948204888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,128,0.030108445220523413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,65536,0.7777617772420248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,128,0.024362666739357844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,65536,2.113847096761068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,16384,0.4755191273159451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,51200,1.4772284825642903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,51200,0.6916729079352485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,16384,0.17006666130489775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,12288,0.4542737536960178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,16384,0.5656142234802246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,65536,2.920914755927192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,16384,0.4966915448506673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,12288,0.1312346723344591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,12288,0.39296711815728086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,65536,2.2376391092936196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,10240,0.33206044303046334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,51200,2.525298648410373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,10240,0.11183999644385444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,10240,0.34697866439819336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,12288,0.3725893232557509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,51200,1.6988835864596898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,7168,0.22715910275777182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,8192,0.09002221955193414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,8192,0.28037601047092015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,10240,0.3039155536227756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,6144,0.2027439938651191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,7168,0.08004711071650188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,7168,0.23689156108432344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,8192,0.2514257695939806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,5120,0.1669173373116387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,6144,0.07069422139061822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,4096,0.1326648924085829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,5120,0.06104533539877998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,6144,0.18735733297136095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,5120,0.16210844781663683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,8192,0.28016800350613064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,4096,0.1373013390435113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,3584,0.10574044121636285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,4096,0.1331768830617269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,3584,0.0458133320013682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,3072,0.09807733032438491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,3584,0.11589954959021674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,3584,0.1186479992336697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,3072,0.04070577687687344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,3072,0.10231288936403062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,2560,0.08249333169725206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,6144,0.1999208927154541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,2048,0.060192889637417264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,2560,0.035863998863432146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,5120,0.1644079950120714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,2560,0.08807555834452312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,1536,0.047367999951044716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,2048,0.031130668189790513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,2048,0.07294044229719374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,2560,0.0912791093190511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,1024,0.03228977653715346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,1536,0.026093333959579468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,2048,0.0762995547718472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,1024,0.02145333256986406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,1536,0.06213866339789497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,768,0.02495199938615163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,1024,0.04920355478922526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,7168,0.23931201299031576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,768,0.03601510988341437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,512,0.01919555498494042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,768,0.019653333557976615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,3072,0.10605155759387547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,256,0.014471999473041959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,512,0.0181048893266254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,512,0.029312888781229656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,4096,0.050776001479890615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,128,0.012039999994966718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,512,0.03496711121665107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,256,0.016398222910033334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,256,0.025447999437650044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,64,0.010465777582592435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,1536,0.057943112320370145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,128,0.016357334123717416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,32,0.011306666665607028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,256,0.028192000256644353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,128,0.02455555564827389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,1024,0.04428799947102865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,64,0.024384000235133704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,128,0.02096533278624217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,32,0.024367999699380662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,768,0.04153777824507819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,65536,0.6604693200853136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,51200,1.4549680285983617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,65536,1.8749947018093531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,51200,0.5466248724195693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,16384,0.4437519974178738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,51200,2.1792577107747397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,16384,0.1511697769165039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,16384,0.4885413381788466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,65536,2.572530746459961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,16384,0.49981599383884007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,12288,0.3963760005103217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,65536,2.053081724378798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,51200,1.5481706195407445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,12288,0.11576178338792588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,12288,0.3684071169959174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,10240,0.28645777702331543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,8192,0.2382417784796821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,10240,0.1042871144082811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,12288,0.3346960014767117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,10240,0.33341066042582196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,8192,0.07990844382180108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,7168,0.21225066979726157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,8192,0.2320622338189019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,10240,0.2860835658179389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,6144,0.17857510513729521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,7168,0.07152533531188965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,7168,0.202984889348348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,8192,0.2428764502207438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,6144,0.0638399985101488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,6144,0.18671466244591606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,4096,0.11869866318172878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,7168,0.20025244024064806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,5120,0.05556355582343208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,5120,0.1477582189771864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,6144,0.1740551127327813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,3584,0.09302755859163071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,4096,0.04532266656557719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,5120,0.14793866210513645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,4096,0.12065066231621636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,3072,0.08709422085020278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,3584,0.041722665230433144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,3584,0.104184885819753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,2560,0.0687697794702318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,4096,0.12219466103447808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,3072,0.03673511081271701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,3072,0.09160444471571182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,2048,0.061385777261522084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,2560,0.03257511059443156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,2560,0.07889155546824138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,3072,0.0957440005408393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,1536,0.04032266802257962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,2048,0.028397334946526423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,2048,0.06554400258594088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,1024,0.02800977892345852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,2560,0.08452533351050483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,5120,0.14338400628831652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,1536,0.023407111565272015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,1536,0.05196266704135471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,768,0.023023999399609033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,1024,0.020336889558368258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,1024,0.03987911012437608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,1536,0.05728977918624878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,512,0.016751110553741455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,768,0.01831733352608151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,768,0.03286400106218126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,1024,0.045943111181259155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,256,0.017651556266678702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,512,0.016537777251667447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,512,0.026467555099063452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,128,0.01513155632548862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,256,0.015296889675988091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,256,0.023232888844278123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,512,0.032607999112870954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,3584,0.11052266756693523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,64,0.013725333743625216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,32,0.014129777749379476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,128,0.01586577792962392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,256,0.024638222323523626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,128,0.022055110997623865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,64,0.02200177808602651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,32,0.021976000732845728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,2048,0.070378667778439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,768,0.03827288746833801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,65536,0.6127902136908637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,51200,1.0807422002156575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,65536,1.5820835961235895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,128,0.01907999979125129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,16384,0.37026132477654355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,51200,0.4843760066562229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,51200,1.557564417521159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,16384,0.13741689258151585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,12288,0.30134667290581596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,65536,2.1426639556884766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,16384,0.5688497755262587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,16384,0.3772871229383681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,65536,1.751574198404948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,12288,0.1252595583597819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,10240,0.23775378863016763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,12288,0.3109831015268962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,8192,0.21230222119225395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,10240,0.10152622063954671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,12288,0.2852862146165636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,10240,0.2564542293548584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,51200,1.3265582190619576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,8192,0.07973866992526583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,7168,0.16624711619483099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,8192,0.20234399371676973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,10240,0.23920533392164442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,6144,0.13370488749610052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,7168,0.06647555695639716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,7168,0.17760888735453287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,8192,0.22365334298875597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,5120,0.12337956163618301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,6144,0.06154133213890923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,7168,0.18875466452704537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,6144,0.15229245026906332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,4096,0.09436177545123631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,5120,0.04909066690338982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,5120,0.1291146675745646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,6144,0.150419553120931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,3584,0.07808888620800443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,4096,0.041101333167817854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,4096,0.10349422030978733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,3072,0.06540711058510675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,5120,0.12829689184824625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,3584,0.035792890522215105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,3584,0.09238755702972412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,4096,0.1058924463060167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,3072,0.03225422236654494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,3072,0.08033066987991333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,3584,0.09394577476713394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,2048,0.048660443888770215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,2560,0.02914666798379686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,2560,0.06969688998328315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,3072,0.08505688773261176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,1536,0.03360622127850851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,2048,0.025250666671329077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,2048,0.05777511331770155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,1024,0.02500533395343357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,1536,0.021254221598307293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,1536,0.04648177822430929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,2048,0.06138844622506035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,768,0.01899111105336083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,1536,0.05098844567934672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,1024,0.0354728897412618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,512,0.0148053334818946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,768,0.016593777471118502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,768,0.029059555795457628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,1024,0.04019733270009359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,256,0.01089422239197625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,2560,0.057406220171186656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,512,0.023464888334274292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,512,0.015178667174445258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,768,0.03405688868628608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,128,0.010058666268984476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,256,0.014094221923086377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,256,0.02055199940999349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,512,0.0288355549176534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,64,0.008714666797055138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,128,0.014041778114106921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,128,0.01993955506218804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,32,0.008819555242856344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,2560,0.07312977976269193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,128,0.01778133379088508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,64,0.019588443968031142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,32,0.01955911186006334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,1024,0.017904000149832833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,65536,0.5478657616509331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,65536,1.180331548055013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,256,0.022145777940750122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,51200,1.124884393480089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,16384,0.3261466556125217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,51200,0.43814754486083984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,51200,1.3550568686591253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,16384,0.344138675265842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,12288,0.29456888304816353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,16384,0.13008888562520346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,65536,1.8788933224148219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,51200,1.1461413701375325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,65536,1.4860124588012695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,10240,0.20239555835723877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,12288,0.10529955228169759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,16384,0.3629777696397569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,12288,0.2719759941101074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,10240,0.0875893301433987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,10240,0.22958578003777397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,12288,0.2518257829878065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,8192,0.06328622500101726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,7168,0.15955289204915366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,8192,0.2059875594245063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,6144,0.1251315540737576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,10240,0.24896266725328234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,7168,0.05676888757281833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,8192,0.17527643839518228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,7168,0.15716889169481066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,5120,0.11374221907721625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,6144,0.049152000082863696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,6144,0.13403466012742785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,4096,0.0829111138979594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,7168,0.15291466977861193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,5120,0.041562666495641075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,5120,0.11182310846116807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,3584,0.07399288813273112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,6144,0.13780177964104548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,4096,0.034563554657830134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,5120,0.11577955881754558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,4096,0.08971199724409316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,3072,0.06588266955481635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,3584,0.03162489003605313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,8192,0.1609075599246555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,4096,0.09411111142900254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,3584,0.0790897806485494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,2560,0.053307556443744235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,3072,0.028432889117134943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,3072,0.06953244739108615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,3584,0.0857235524389479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,2048,0.043192889955308705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,2560,0.02556177808178796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,2560,0.059771557648976646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,3072,0.07636088795132108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,2048,0.022058667408095464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,2048,0.04975110954708523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,2560,0.06611733304129706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,1024,0.024697777297761705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,1536,0.01865511139233907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,1536,0.03990755478541056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,2048,0.056200000974867076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,768,0.019501333435376484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,1024,0.01588800052801768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,1024,0.03012088934580485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,1536,0.04672266708479988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,512,0.014854222536087036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,768,0.015134221977657743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,768,0.025591999292373657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,1024,0.03637689020898607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,256,0.011521778172916837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,512,0.013914666242069669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,256,0.01798488861984677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,512,0.02556266718440586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,256,0.013078221844302284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,128,0.008823110825485652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,256,0.020918221937285528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,1536,0.034020443757375084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,128,0.013021333350075616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,64,0.00777333312564426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,32,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,128,0.01723377737734053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,128,0.016930666234758165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,64,0.017338666650984023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,32,0.017183999220530193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,512,0.02064622276359134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,65536,0.42652710278828937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,65536,0.9828151067097982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,16384,0.27292532391018337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,51200,0.33263643582661945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,51200,1.0816924836900499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,65536,1.5581049389309352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,16384,0.10207733180787827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,16384,0.31722667482164174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,12288,0.2073742283715142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,65536,1.295938703748915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,768,0.03170488940344916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,16384,0.3049351109398736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,12288,0.08503911230299209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,51200,0.9370159573025174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,10240,0.17497421635521782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,51200,0.7393591139051648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,12288,0.21533511744605172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,8192,0.1385342280069987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,10240,0.06958311133914523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,7168,0.12302133772108291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,10240,0.17780889405144584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,12288,0.22596799002753365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,8192,0.15914400418599448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,6144,0.10084266795052423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,7168,0.047641777329974704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,10240,0.18997511598798963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,7168,0.12920799520280626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,8192,0.15437066555023193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,6144,0.04120533333884345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,6144,0.11307911078135173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,7168,0.15132178200615778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,4096,0.06520622306399874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,5120,0.035440001222822405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,5120,0.0934248897764418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,6144,0.11995911598205566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,3584,0.05560266971588135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,4096,0.030047112041049536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,4096,0.07631111145019531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,5120,0.10105511214998032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,3072,0.04678577846950955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,3584,0.06683378087149726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,4096,0.08410133255852593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,2560,0.040463109811147056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,8192,0.053764445914162524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,3072,0.025215999947653875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,3584,0.07642577754126655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,3072,0.05834488736258613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,2048,0.03404533200793796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,2560,0.02255200015174018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,3072,0.06819911135567559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,2560,0.050379554430643715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,1536,0.024995555480321247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,5120,0.08325777451197307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,2048,0.04270577761862013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,2048,0.0199991116921107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,2560,0.058295110861460366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,1024,0.019017777509159513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,1536,0.01756977703836229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,1536,0.03383200036154853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,2048,0.050352887974845044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,768,0.015263999501864115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,1024,0.015167999598715039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,1024,0.02569155560599433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,512,0.01146133326821857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,1536,0.04127555423312717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,1024,0.03271999955177307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,768,0.021919111410776775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,3584,0.02733777794573042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,256,0.009728000395827824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,512,0.01755911111831665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,768,0.026895110805829365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,128,0.008968888885445064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,512,0.023023999399609033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,256,0.015536889433860779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,256,0.012054222325483957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,64,0.008571555217107138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,256,0.019519999623298645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,128,0.012083555261294046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,32,0.008943999807039896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,64,0.0148026661740409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,128,0.015796444482273526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,32,0.014774221513006421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,768,0.013797333670987023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,65536,0.3838506539662679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,512,0.01275644451379776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,65536,0.9007387161254883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,51200,0.7520248625013563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,16384,0.25221067004733616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,128,0.01479288935661316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,51200,0.3180026743147108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,65536,1.2518000072903104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,16384,0.09523910946316189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,12288,0.19229511419932047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,65536,1.1144800186157227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,51200,0.9882035785251193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,12288,0.07532622416814168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,10240,0.16824711693657768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,12288,0.19677866829766166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,8192,0.11667377418941921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,51200,0.8732257948981391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,10240,0.06951910919613309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,12288,0.19832444190979004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,10240,0.16698755158318412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,7168,0.109860446718004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,8192,0.05533599853515625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,10240,0.1801582177480062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,8192,0.12917510668436685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,6144,0.0939084423912896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,7168,0.04309422108862135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,7168,0.11843021710713704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,8192,0.14841689003838435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,5120,0.07838844590716891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,6144,0.038198222716649376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,16384,0.28385599454243976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,7168,0.12104888757069905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,6144,0.0979422198401557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,4096,0.06368177466922335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,5120,0.03244977858331468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,6144,0.1074133316675822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,3584,0.055762668450673424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,4096,0.02715911136733161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,16384,0.26974132325914174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,4096,0.06738222307629056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,3072,0.04650311006440056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,5120,0.09150489171346028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,3584,0.024736000431908503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,2560,0.03982399900754293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,3584,0.06091377470228407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,3072,0.02256444427702162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,3072,0.053391999668545194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,3584,0.06841244300206502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,2560,0.020578665865792167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,3072,0.06159466505050659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,2560,0.04639200038380093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,1536,0.025701334079106648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,2048,0.01792888840039571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,2560,0.05344266692797343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,2048,0.03894844320085313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,1024,0.01999200052685208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,1536,0.015782222151756287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,1536,0.031238221459918555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,5120,0.085116446018219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,1024,0.01352533366945055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,1024,0.023725334140989516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,1536,0.037744889656702675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,768,0.016241777274343703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,512,0.013462222284740873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,768,0.01308444473478529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,1024,0.02886755598915948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,4096,0.0759377810690138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,768,0.01999022232161628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,256,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,768,0.024901333782407973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,2048,0.033422221740086876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,512,0.01202933324707879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,512,0.016209777858522203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,128,0.006893333461549547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,256,0.014161777165200977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,256,0.011359111302428775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,64,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,128,0.013820444544156393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,32,0.008283555507659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,256,0.0177208897140291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,2048,0.04571911030345493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,128,0.014819555812411837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,64,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,32,0.013637333280510373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,65536,0.3212089008755154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,65536,0.6130880249871148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,51200,0.4841368993123372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,512,0.021292444732454088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,128,0.011208888557222156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,16384,0.1856151156955295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,65536,1.0658719804551866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,65536,0.9314639833238391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,16384,0.07799022065268622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,12288,0.13991733392079672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,16384,0.2245182196299235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,51200,0.851155545976427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,10240,0.1264266702863905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,51200,0.7372328970167371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,12288,0.06738311052322388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,16384,0.22813243336147734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,12288,0.1727679967880249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,8192,0.09474222527609931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,10240,0.0595760014322069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,10240,0.15302133560180664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,12288,0.17320711082882354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,7168,0.08721866872575547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,8192,0.04281511240535312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,8192,0.11510133743286133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,6144,0.07184088892406888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,10240,0.14630666044023302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,7168,0.03707911239729987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,8192,0.12101956208546956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,51200,0.24529155095418295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,5120,0.06262666649288602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,6144,0.03203466534614563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,6144,0.08808444605933295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,7168,0.10724266370137532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,5120,0.029338667790095013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,5120,0.07466666566001044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,6144,0.0936711099412706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,3584,0.03995199998219808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,4096,0.02506133251720005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,5120,0.08103377951516046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,4096,0.06058310800128513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,3072,0.03341777788268195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,3584,0.02199911077817281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,3584,0.05401600069469876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,4096,0.0669671098391215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,2560,0.031360887818866305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,3072,0.020346666375796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,3072,0.04804799954096476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,3584,0.06016888883378771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,2048,0.02453688946035173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,2560,0.018593778212865193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,2560,0.041654222541385226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,3072,0.05445422066582573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,1536,0.019711999429596793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,2048,0.01665866706106398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,7168,0.10328266355726455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,2048,0.0350471105840471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,1024,0.014041778114106921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,1536,0.01479555500878228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,1536,0.028318223026063707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,2048,0.040258665879567467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,768,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,1024,0.021402666966120403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,1024,0.012744888663291931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,1536,0.033358222908443876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,512,0.00865955568022198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,1024,0.02574666672282749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,768,0.018124444617165458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,256,0.006451555424266391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,768,0.02312622302108341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,512,0.014888889259762235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,512,0.019263111882739596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,256,0.010399999717871347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,256,0.01311911145846049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,128,0.006024000131421619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,128,0.010656889114114972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,64,0.005504888792832692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,256,0.01702133317788442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,128,0.012441777520709567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,32,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,2560,0.04629155662324694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,128,0.014370666609870063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,4096,0.05028533273273044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,64,0.01239733315176434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,32,0.012420444852775998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,768,0.01181688904762268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,65536,0.5671137703789605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,65536,0.3012648953331841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,512,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,51200,0.4227120081583659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,16384,0.16467021571265325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,51200,0.23868799209594727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,65536,0.8496168984307183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,16384,0.19603199428982207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,12288,0.12951378027598062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,16384,0.07493778069814046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,51200,0.6621902253892686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,65536,0.8779733445909289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,10240,0.1188124418258667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,16384,0.22397777769300672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,12288,0.05836355686187744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,51200,0.7007493442959256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,8192,0.09347555372450088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,10240,0.05013777812321981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,10240,0.131205333603753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,7168,0.08016800218158297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,8192,0.0385484430525038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,8192,0.1043048898379008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,10240,0.15140533447265625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,7168,0.034563554657830134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,6144,0.06429777542750041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,8192,0.11817422178056504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,7168,0.09029599693086411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,5120,0.05524444580078125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,6144,0.03141155507829454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,7168,0.10423022508621216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,6144,0.0785742203394572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,4096,0.042740444342295326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,5120,0.027497778336207073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,5120,0.06588355700174968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,6144,0.09138400024837917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,4096,0.0232924438185162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,5120,0.07804444763395521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,4096,0.05242933498488533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,12288,0.1481484439637926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,3072,0.034479111433029175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,3584,0.046464890241622925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,4096,0.06500977940029569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,2560,0.028127110666698877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,12288,0.1736462248696221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,3072,0.019319110446506076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,3584,0.05927377939224243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,2048,0.02408355474472046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,3072,0.05315466721852621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,2560,0.03583822316593594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,2048,0.015842666228612263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,2048,0.03067022230890062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,1536,0.018969777557584975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,2048,0.039798223310046725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,3584,0.03753333290417989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,1024,0.014475555883513557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,1536,0.01422133379512363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,1536,0.03126400046878391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,3584,0.02111377815405528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,1024,0.018591110905011494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,768,0.011043555206722684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,1024,0.012440000143316058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,3072,0.04151111178927951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,512,0.008943110704421997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,768,0.011635555161370171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,1024,0.025352888637118872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,768,0.01588088936275906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,2560,0.01754044493039449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,256,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,512,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,768,0.021907554732428655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,512,0.013120888835854001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,128,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,2560,0.04599466588762072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,256,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,256,0.011713777979214987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,64,0.005758222192525864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,512,0.018589332699775696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,32,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,128,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,128,0.011359111302428775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,1536,0.02436622149414486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,64,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,128,0.012432888978057437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,32,0.011192888849311404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,65536,0.47240177790323895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,65536,0.2606728871663411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,51200,0.3608728779686822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,16384,0.1305262247721354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,51200,0.21946844789716932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,65536,0.7025618023342557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,51200,0.548201772901747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,16384,0.17043733596801758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,16384,0.06782311201095581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,65536,0.7509528795878092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,51200,0.5730071067810059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,256,0.015815999772813585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,12288,0.05195288856824239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,10240,0.09261422024832831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,12288,0.1271644433339437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,8192,0.0684551133049859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,10240,0.04766222172313266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,12288,0.14761866463555232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,10240,0.10668355226516724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,7168,0.0602933300866021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,8192,0.03359288970629374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,8192,0.08591110838784112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,10240,0.12467643949720596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,6144,0.05195200112130907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,7168,0.030960887670516968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,8192,0.10108266936408149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,7168,0.07607911030451457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,5120,0.040503998597462974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,12288,0.10028000010384454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,6144,0.06690933307011922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,6144,0.02870133188035753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,7168,0.089793774816725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,4096,0.03466488917668661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,5120,0.024475556280877855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,5120,0.05561955769856771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,6144,0.07826755444208781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,16384,0.18992443879445395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,4096,0.02009600069787767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,5120,0.06785955693986681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,3072,0.026036444637510512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,3584,0.018467555443445843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,4096,0.05646489063898722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,3584,0.041577776273091636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,2560,0.022840888963805303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,3584,0.0516480008761088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,3072,0.036877334117889404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,2560,0.03174488743146261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,2560,0.015463999576038785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,2048,0.018377777602937486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,3072,0.04564977685610453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,1536,0.013518222504191928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,2048,0.01401422255569034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,2048,0.02721422248416477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,3584,0.029519110918045044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,1536,0.02204622162712945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,1536,0.012839111189047495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,1024,0.010484444598356882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,4096,0.04612088865704007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,768,0.0087013335691558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,1536,0.027962668074501887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,1024,0.01111644423670239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,1024,0.016582222448454965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,512,0.006760889043410619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,768,0.010628444453080496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,768,0.014339556296666464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,3072,0.016916443904240925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,256,0.0058497778243488735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,512,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,768,0.01944533321592543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,512,0.011988444460762871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,128,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,256,0.009407999614874521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,256,0.010400888820489248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,512,0.016840888394249808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,64,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,128,0.009495110975371467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,32,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,256,0.01495911180973053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,2048,0.033915556139416166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,64,0.009727111293209923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,128,0.012094222009181976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,32,0.009995555712117089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,65536,0.3608986536661784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,65536,0.19350755214691162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,1024,0.022341334157519873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,51200,0.269524441825019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,16384,0.10376177893744574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,65536,0.5463297632005479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,51200,0.1532720062467787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,51200,0.45370843675401473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,2560,0.03958755400445726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,65536,0.587897777557373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,16384,0.1430222193400065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,12288,0.07967022392484877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,128,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,10240,0.06324711110856798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,12288,0.040604445669386126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,16384,0.1552693314022488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,51200,0.4618808958265517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,12288,0.10552710956997341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,8192,0.050380445188946195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,10240,0.08836978011661106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,10240,0.03513066636191474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,7168,0.04320266511705187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,12288,0.1196097797817654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,8192,0.026258667310078938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,6144,0.037040889263153076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,10240,0.10247911347283258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,7168,0.02381955583890279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,7168,0.06334311432308622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,5120,0.029158221350775823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,6144,0.02128177715672387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,6144,0.054818666643566556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,7168,0.07255378034379747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,6144,0.06498222218619452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,5120,0.01868799991077847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,16384,0.050460444556342236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,4096,0.02479377720091078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,5120,0.0462097790506151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,3584,0.02165777815712823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,4096,0.016530666086408827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,5120,0.0561351113849216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,4096,0.03856089048915439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,3072,0.019693333241674636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,3584,0.03470844361517165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,3584,0.01516266663869222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,4096,0.04706844356324938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,2560,0.016433777080641854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,3072,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,8192,0.07157599925994873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,3072,0.030763556559880573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,3584,0.04236533244450887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,2048,0.013787555197874704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,2560,0.012874667015340594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,2560,0.02683111031850179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,1536,0.011121778024567498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,3072,0.037441776858435735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,2048,0.01184177812602785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,2048,0.02275200022591485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,1024,0.008625778059164682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,2560,0.03331911232736375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,1536,0.01074577785200543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,1536,0.018583999739752877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,768,0.007432888779375289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,1024,0.014448000325096978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,1536,0.023573334018389385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,1024,0.0192577772670322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,768,0.012432888978057437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,768,0.009673777553770278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,512,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,512,0.009024889104896123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,256,0.004569777597983678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,768,0.01706044375896454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,512,0.010411555568377176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,128,0.0038657776183552216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,256,0.008771555291281806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,256,0.009132444030708736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,128,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,256,0.013461333182122974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,128,0.00871733327706655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,64,0.0035422220826148987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,32,0.003859555555714501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,2048,0.02848177817132738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,128,0.010581333604123859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,1024,0.010031110710567897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,64,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,65536,0.25261423322889537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,51200,0.22148088614145914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,65536,0.1794604460398356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,65536,0.44588624106513125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,16384,0.0699768861134847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,51200,0.35039199723137754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,51200,0.142593781153361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,512,0.015199111567603217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,12288,0.0539226664437188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,16384,0.04879466692606608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,65536,0.5745404561360677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,16384,0.11520444022284614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,10240,0.04458399944835239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,12288,0.038725333081351385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,12288,0.086681776576572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,51200,0.45398489634195965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,16384,0.15397866566975912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,8192,0.03506133291456435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,32,0.008734222087595198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,10240,0.033407111962636314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,12288,0.11838044060601129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,7168,0.03089955449104309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,8192,0.025275554921891954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,8192,0.05890044238832262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,6144,0.0262071109480328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,10240,0.10082044204076131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,7168,0.023019555542204116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,7168,0.052654223309622876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,5120,0.022776888476477727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,6144,0.020555555820465088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,8192,0.0840853320227729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,7168,0.07271022266811795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,6144,0.04564177658822802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,4096,0.018664888209766813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,5120,0.018195556269751657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,6144,0.06410044431686401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,3584,0.01684177749686771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,4096,0.032590223683251276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,4096,0.0158915552828047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,5120,0.05499910977151659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,3072,0.01475288967291514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,3584,0.01482311056719886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,4096,0.0459591117170122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,3584,0.029233776860766943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,2560,0.01240355521440506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,10240,0.07260266939798991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,3072,0.013804444836245643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,3072,0.025622222158643935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,3584,0.04145688811937968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,2048,0.010976889067225985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,2560,0.01278133359220293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,2560,0.022463111413849726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,3072,0.03664266732003953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,8192,0.08164178000556098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,1536,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,2048,0.019293333093325298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,2560,0.032186667124430336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,1024,0.006779555645253923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,1536,0.016173332929611206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,1536,0.01072799993885888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,5120,0.03919911053445604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,768,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,1536,0.02312799957063463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,1024,0.01000355515215132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,768,0.00941955546538035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,512,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,768,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,1024,0.018938667244381376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,768,0.01686311099264357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,512,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,256,0.004005333201752769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,512,0.009335111412737105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,256,0.008625778059164682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,128,0.0035146665241983202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,256,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,2048,0.011725333001878528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,256,0.01309511148267322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,64,0.003524444583389494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,128,0.008340444829728868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,128,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,32,0.0034959999223550162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,1024,0.012472000386979846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,128,0.01072799993885888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,64,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,65536,0.22487910588582358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,32,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,51200,0.1746844450632731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,65536,0.17605422602759468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,16384,0.054471111959881253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,65536,0.36523644129435223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,51200,0.14720800187852648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,51200,0.28991378678215873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,512,0.014825777875052558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,12288,0.03969600134425693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,16384,0.051022221644719444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,65536,0.43760355313618976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,16384,0.09603555334938897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,10240,0.034556445148256086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,51200,0.3449706766340468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,12288,0.03862311111556159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,12288,0.0727528863483005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,16384,0.11986755000220405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,2048,0.027435556054115295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,8192,0.026434666580624048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,7168,0.02346222268210517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,10240,0.03369866808255514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,10240,0.06079466475380791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,12288,0.09230311049355401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,8192,0.024877332978778418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,6144,0.019318222999572754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,7168,0.022504000200165644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,10240,0.07857866419686212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,5120,0.016145777371194627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,6144,0.020233778489960563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,8192,0.06403199831644694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,6144,0.03853244582811991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,7168,0.05679021941290962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,6144,0.05050577720006307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,5120,0.01792355544037289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,5120,0.03292799989382426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,3584,0.011938666303952536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,4096,0.01587022178702884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,4096,0.02718577782313029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,5120,0.042950222889582314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,3072,0.010755555497275459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,3584,0.024744888146718342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,4096,0.03644711110326979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,3072,0.022086222966512043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,3584,0.033018665181265935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,2560,0.00943022221326828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,8192,0.04939377639028761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,3072,0.028772443532943726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,7168,0.045111109813054405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,2048,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,2560,0.019260444574885897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,2560,0.012780444489585029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,1536,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,4096,0.013346666263209449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,2560,0.026064000195927087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,2048,0.011681777735551199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,1024,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,2048,0.016563556260532804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,1536,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,1536,0.013880000346236758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,2048,0.022672888305452134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,768,0.004592000196377437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,1024,0.009919111099508073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,1024,0.01075111081202825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,1536,0.019193778435389202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,512,0.0038817777401871155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,768,0.009353777600659264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,1024,0.015848888291252982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,768,0.009710222482681274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,256,0.0032266666077905228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,3072,0.013750221994188098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,512,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,512,0.00852177788813909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,128,0.00282577777074443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,256,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,512,0.01315111087428199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,256,0.007246221933099959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,64,0.0028026666906144884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,128,0.006911110960774952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,32,0.0028204443968004654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,256,0.012048888537618848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,128,0.008376888930797577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,128,0.009315555294354757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,64,0.006881777611043718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,32,0.006684444430801604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,51200,0.12949332926008436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,3584,0.014829332629839579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,65536,0.13253511322869196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,65536,0.3174320062001546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,16384,0.043303999635908336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,51200,0.10623288816875881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,768,0.014499555031458536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,51200,0.2501280042860243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,16384,0.08433422115113999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,65536,0.42688711484273273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,12288,0.03188800149493747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,16384,0.03627644313706292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,10240,0.025998221503363714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,12288,0.02882222334543864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,51200,0.33870577812194824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,16384,0.11869245105319554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,12288,0.06388888756434123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,8192,0.019183110859658983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,10240,0.02570044497648875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,10240,0.053527110152774386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,7168,0.01718844473361969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,12288,0.09042133225335015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,8192,0.018566222654448617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,8192,0.04360533422893948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,10240,0.07624089055591159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,6144,0.015274667077594332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,7168,0.016920000314712524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,7168,0.03912711143493652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,5120,0.01311288856797748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,8192,0.0627066691716512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,6144,0.015497777197096082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,7168,0.055833776791890465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,4096,0.011077333655622272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,5120,0.014170666535695394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,5120,0.029300444655948218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,6144,0.049513777097066246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,4096,0.012429333395428128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,4096,0.024096000525686476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,5120,0.0425546665986379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,3072,0.009189333352777693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,3584,0.02206755512290531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,3584,0.011756444142924415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,65536,0.15402844217088488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,2560,0.008029333419269985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,3072,0.011015999648306104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,3584,0.03194311261177063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,2560,0.010633777413103314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,2048,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,3072,0.028394665982988145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,2560,0.017536888519922893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,2048,0.010013333625263637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,1536,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,2048,0.015054222610261707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,3584,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,2048,0.02221511138810052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,1024,0.004800888813204235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,1536,0.012684444586435953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,1536,0.009400889277458191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,4096,0.03519999980926514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,1024,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,1536,0.018587556150224473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,768,0.0041804446114434134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,1024,0.01000088867213991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,512,0.003536000019974179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,3072,0.0199946661790212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,768,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,1024,0.015505777464972602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,256,0.003220444545149803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,768,0.009004444711738164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,512,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,512,0.007725333174069722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,128,0.003156444471743372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,768,0.013981334037250943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,256,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,256,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,512,0.012439111040698158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,2560,0.025199999411900837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,64,0.0028284444577164124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,6144,0.03439377744992574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,32,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,128,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,256,0.011092444260915121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,128,0.006144888699054718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,65536,0.09772355688942803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,64,0.00611200018061532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,51200,0.07995200157165527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,128,0.008674666285514832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,32,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,16384,0.029640889830059473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,65536,0.13061155213250056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,65536,0.27477065722147626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,51200,0.10519199901156956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,12288,0.024969778127140466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,51200,0.21533332930670845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,16384,0.07238311237759061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,16384,0.03681066632270813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,10240,0.020840888222058613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,65536,0.41416533788045246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,12288,0.028357333607143823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,51200,0.3259013228946262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,12288,0.0551111102104187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,8192,0.01759111053413815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,10240,0.0248346659872267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,10240,0.046441776884926685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,7168,0.014428445034556918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,12288,0.08752622207005818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,8192,0.038104888465669416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,10240,0.07392799854278564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,6144,0.013146667016877068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,7168,0.016872889465755887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,7168,0.03409333361519708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,8192,0.06073333157433403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,5120,0.010393777655230628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,7168,0.053187555736965604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,6144,0.0155102229780621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,6144,0.030057777961095173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,4096,0.008197333249780867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,5120,0.02572000026702881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,5120,0.013861333330472311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,3584,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,6144,0.04750400119357639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,4096,0.012446222205956778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,5120,0.039932443035973444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,4096,0.021410667233996924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,3072,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,16384,0.1139262252383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,3584,0.011770666473441653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,4096,0.0336773329310947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,3584,0.019440889358520508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,2560,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,3072,0.011168888873524137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,3072,0.01757066614098019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,2048,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,2560,0.010676444404655032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,8192,0.018200889229774475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,2560,0.01553422212600708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,1536,0.004593777573770947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,3072,0.027276444766256545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,2048,0.010192888478438059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,2048,0.013459555804729462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,2560,0.024098666177855596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,1536,0.009369778136412302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,2048,0.021352888809310064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,1536,0.011368888947698804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,768,0.003318222239613533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,1024,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,1024,0.009197333620654212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,1536,0.018231999542978074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,512,0.002874666617976295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,768,0.008358221915033128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,768,0.00811644477976693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,1024,0.01483022173245748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,512,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,768,0.013764444324705334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,512,0.012105777859687805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,256,0.005976000179847081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,128,0.002455111179086897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,3584,0.030800888935724895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,256,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,64,0.0025031111306614345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,128,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,32,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,128,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,128,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,65536,0.07298666901058622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,64,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,32,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,51200,0.058555556668175586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,65536,0.12997599442799887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,256,0.0027422221998373666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,65536,0.2364711231655545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,16384,0.021934222843911912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,512,0.007059555914666917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,51200,0.10445955726835464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,256,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,12288,0.013737777868906656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,65536,0.4091226789686415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,16384,0.06212800078921848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,16384,0.03664088911480374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,16384,0.11315200063917373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,51200,0.3227439986334907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,1024,0.0039057777159743835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,10240,0.011959111524952782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,12288,0.04760444495413038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,8192,0.012986666626400419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,10240,0.024823110964563157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,12288,0.0864400002691481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,10240,0.04033689035309686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,7168,0.01130222198035982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,8192,0.03302399979697333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,8192,0.018945778409639995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,10240,0.07299111286799113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,6144,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,7168,0.01716800034046173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,8192,0.05994844436645508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,51200,0.18552622530195448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,6144,0.025977777110205755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,5120,0.00904799997806549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,6144,0.015678221980730694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,7168,0.053343110614352755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,4096,0.007786666353543599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,5120,0.014141332772043018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,6144,0.04580977890226576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,12288,0.02871555421087477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,3584,0.007094221810499827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,4096,0.012490666574902005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,4096,0.01891999940077464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,5120,0.039071109559800886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,3072,0.0063662222690052455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,3584,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,3584,0.017232000827789307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,4096,0.03336888882848952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,2560,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,3072,0.015615999698638916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,3072,0.011095999843544431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,3584,0.030245333909988403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,2048,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,7168,0.0295413335164388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,2560,0.01384800010257297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,3072,0.02709688908523983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,1536,0.004233777936961916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,2048,0.010048888623714447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,2048,0.012102222277058495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,2560,0.02402488887310028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,1024,0.003543111185232798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,2048,0.021191999316215515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,1536,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,1536,0.01051288925939136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,5120,0.02257777750492096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,1024,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,1024,0.008522666990756989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,1536,0.017874666386180453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,768,0.008316444853941599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,768,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,1024,0.014902222487661572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,256,0.00249422217408816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,768,0.013094222380055321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,512,0.006685333533419504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,256,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,256,0.005655111124118169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,128,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,512,0.011744889120260874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,2560,0.01071466671095954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,64,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,128,0.007664889097213745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,256,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,32,0.002242666565709644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,128,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,65536,0.06803644365734525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,64,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,51200,0.053344001372655235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,768,0.003177777760558658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,32,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,16384,0.019015999303923715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,12288,0.015560888581805758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,10240,0.013807999591032663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,512,0.0028320000403457214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,65536,0.23632356855604383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,8192,0.011756444142924415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,7168,0.011039111349317761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,51200,0.18544533517625597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,16384,0.06203022268083361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,6144,0.009761778016885122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,5120,0.00833511104186376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,512,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,4096,0.007196444604131911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,7168,0.028856890069113836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,6144,0.025751999682850305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,5120,0.02230311102337307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,3072,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,4096,0.01888177792231242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,2560,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,3584,0.01718488832314809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,2048,0.004523555437723796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,3072,0.015684444043371413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,1536,0.0038977778620190094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,2560,0.013787555197874704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,2048,0.012088889049159156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,1024,0.003289777785539627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,1536,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,12288,0.047671112749311656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,768,0.003148444410827425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,512,0.0026977778308921387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,10240,0.04018933243221707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,1024,0.008536888493431939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,256,0.002512000087234709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,128,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,8192,0.032602667808532715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,64,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,768,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,3584,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,32,0.0021608887861172357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,512,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,256,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,65536,0.06628977590137057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,128,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,51200,0.050056000550587974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,16384,0.016536000702116225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,64,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,12288,0.01034400032626258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,32,0.005231110999981563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,10240,0.009310222334331935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,8192,0.008063111040327284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,65536,0.23623198933071562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,51200,0.18536444505055746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,7168,0.007679111427730984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,12288,0.04757510953479343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,6144,0.007175111108356052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,5120,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,10240,0.03986577855216132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,8192,0.032474666833877563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,4096,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,3584,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,7168,0.028914666838116113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,6144,0.025595555702845257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,3072,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,2560,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,4096,0.018923555811246235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,3584,0.017235555582576327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,2048,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,1536,0.0039022221333450745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,3072,0.015629332926538255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,1024,0.003192000091075897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,2560,0.013787555197874704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,768,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,2048,0.012049777640236748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,1536,0.01036888857682546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,512,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,256,0.0022960000981887183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,1024,0.008352888955010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,128,0.002159111201763153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,768,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,64,0.0018586667461527719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,512,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,16384,0.06191466914282905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,256,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,128,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,64,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,32,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,5120,0.02233244478702545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,32,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,16384,1.213862207200792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,16384,3.4590284559461804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,12288,2.734920925564236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,16384,4.114911821153429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,12288,0.8673057556152344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,12288,2.976572460598416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,51200,4.501092274983724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,10240,0.73199552959866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,10240,2.2111004723442926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,10240,2.370003594292535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,8192,0.6054533322652181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,8192,1.8232417636447484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,7168,1.6071493360731337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,7168,0.5761946572197808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,8192,1.9846568637424047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,6144,0.4621075524224176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,7168,1.5835866928100586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,6144,1.4136186175876195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,5120,1.2713386747572157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,5120,0.3973822328779433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,6144,1.357952011956109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,4096,0.947572496202257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,4096,0.32207200262281627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,5120,1.1959440443250868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,3584,0.2884151140848796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,3584,0.8613235685560438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,4096,0.9290587107340494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,3072,0.6436230871412489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,51200,11.022497389051649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,3584,0.8200293117099338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,2560,0.5921617613898383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,3072,0.6956835322909884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,2048,0.4177102247873942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,2560,0.5924746725294325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,2048,0.18944977389441597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,1536,0.3462515672047933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,2048,0.5271760092841254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,1024,0.23609420988294813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,3072,0.268667565451728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,1536,0.1637048853768243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,1536,0.3671920034620497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,768,0.19011467032962373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,1024,0.12224533822801377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,1024,0.3037244478861491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,768,0.11200800206926133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,768,0.21877511342366537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,256,0.0889893306626214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,2560,0.22372089491950142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,512,0.09711822536256577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,512,0.19011110729641387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,128,0.07958666483561198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,256,0.08488533231947158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,256,0.16808444923824736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,64,0.06615555286407471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,128,0.08312622043821546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,32,0.06637689140107897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,128,0.15946133931477866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,64,0.1616951094733344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,32,0.16422400209638807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,512,0.15903200043572321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,51200,17.703938802083332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,51200,3.496405283610026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,128,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,65536,4.53153313530816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,16384,0.9157999886406792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,16384,2.9925261603461375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,12288,0.7005528873867459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,12288,2.5141448974609375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,16384,3.460465749104818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,51200,8.891571044921875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,12288,2.297515657212999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,10240,0.5820524427625868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,8192,0.4767626656426324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,10240,2.2325288984510636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,8192,1.527451515197754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,10240,1.933066685994466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,8192,1.4079128901163738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,65536,11.274598015679253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,7168,0.41489511066012913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,7168,1.4791884952121312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,6144,1.1701173782348633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,6144,1.0340026219685872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,6144,0.3627271122402615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,7168,1.2635866800944011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,5120,0.3586222330729167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,5120,0.9283840391370984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,4096,0.26041332880655926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,4096,0.6814639833238391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,5120,0.8867191208733453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,4096,0.7108888626098633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,3584,0.23276178042093912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,3584,0.6774702072143555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,3072,0.532367123497857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,3584,0.6782257821824816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,3072,0.5232497851053873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,2560,0.18440800242953828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,51200,13.530511644151476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,2048,0.3411413298712836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,2048,0.15085244178771973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,1536,0.2893768946329753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,2560,0.6602755652533637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,2048,0.36893865797254777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,1024,0.20052711168924967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,1536,0.1209191083908081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,1536,0.3208719889322917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,1024,0.09683999750349258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,768,0.16687822341918945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,1024,0.25512266159057617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,512,0.1156088908513387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,768,0.08751110898123847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,256,0.07338577508926392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,512,0.07746133539411756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,768,0.172705782784356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,128,0.05972177452511258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,256,0.07040711243947347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,512,0.17267289426591662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,256,0.13370933797624376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,64,0.05625333388646444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,128,0.06741066773732503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,32,0.05894311269124349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,128,0.12631555398305258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,32,0.13233333163791233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,64,0.1285368866390652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,3072,0.20380266507466635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,2560,0.4441760116153293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,65536,1.4491528405083551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,16384,1.0489947001139324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,51200,1.0778186586168077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,65536,17.67948744032118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,16384,0.32250844107733834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,12288,0.23040265507168242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,12288,0.8719031016031901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,12288,0.7534569104512533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,16384,1.030268457200792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,51200,3.145465850830078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,10240,0.19746843973795572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,10240,0.6490026579962836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,8192,0.48863734139336484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,65536,4.058508555094401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,8192,0.20525422361161974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,10240,0.6361884540981716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,7168,0.41643110911051434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,8192,0.493053330315484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,7168,0.1842719978756375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,6144,0.35355377197265625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,51200,4.199419657389323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,6144,0.12388088968065049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,7168,0.4281946553124322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,5120,0.31038400861952037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,5120,0.10690489080217148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,4096,0.23503377702501085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,6144,0.38643466101752383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,5120,0.3160906632741292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,4096,0.10980000098546346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,3584,0.2039875586827596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,3584,0.0793164438671536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,4096,0.2340159946017795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,3072,0.16518044471740723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,2560,0.14556088712480333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,3072,0.08053421974182129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,3584,0.26775556140475804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,2048,0.12089510758717854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,2560,0.062182221147749156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,65536,5.554249657524957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,1536,0.0936080018679301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,1024,0.06370044416851468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,2048,0.057032889790005155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,2560,0.15464177396562365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,1536,0.0432951119210985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,768,0.04943289028273689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,2048,0.1251173284318712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,512,0.03639288743336996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,1024,0.03995022177696228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,1536,0.1033431092898051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,256,0.028451555305057104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,768,0.03357244531313578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,512,0.03152799937460158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,128,0.022983999715911016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,1024,0.07712355587217543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,512,0.05219733383920458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,256,0.028685334655973647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,256,0.04731377628114489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,64,0.01994311147265964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,128,0.027292443646325007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,32,0.020607999629444547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,128,0.04452355702718099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,64,0.04512622290187412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,32,0.04581422275967068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,3072,0.2148488892449273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,768,0.06136977672576904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,16384,0.7483653492397733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,65536,1.0093741946750217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,51200,0.8019519911872016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,16384,0.28311112191942006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,16384,0.7318595780266656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,12288,0.6711946593390571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,51200,2.0516346825493708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,12288,0.21527111530303955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,12288,0.8258799976772733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,10240,0.43897154596116805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,10240,0.15174844529893664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,8192,0.34925156169467503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,8192,0.12607022126515707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,10240,0.515485339694553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,7168,0.3084942234887017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,8192,0.3692106670803494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,7168,0.10999911361270481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,6144,0.2763742340935601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,7168,0.350092437532213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,6144,0.09547999832365249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,51200,3.1063609653049045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,5120,0.23119823137919107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,6144,0.28250310156080455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,5120,0.08328000042173597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,4096,0.18352444966634116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,4096,0.0747706691424052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,3584,0.14708977275424534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,5120,0.2520471149020725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,3584,0.0613386697239346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,3072,0.12968977292378744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,4096,0.19095022148556182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,65536,2.6492578718397355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,3584,0.15773955980936685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,2560,0.10739288727442424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,3072,0.054809778928756714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,2560,0.049485332436031766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,2048,0.09411111142900254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,2048,0.04247200157907274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,2560,0.12059911092122395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,65536,4.31052992078993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,1024,0.052367998494042285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,1536,0.03472355670399136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,2048,0.0969582200050354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,768,0.0410497784614563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,512,0.027103111147880554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,1024,0.031529777579837374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,1536,0.07714755667580499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,768,0.026365333133273657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,1024,0.06044977903366089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,256,0.019479999939600628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,768,0.0472924444410536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,512,0.024316444993019104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,128,0.01752800080511305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,256,0.021938666701316833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,512,0.04143911268975999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,64,0.01479288935661316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,32,0.0147724449634552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,128,0.02165599995189243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,128,0.03351022137535943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,64,0.03423110975159539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,32,0.03365066647529602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,1536,0.06508444415198432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,256,0.035383111900753446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,65536,0.9123813841078016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,3072,0.138044450018141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,51200,0.7090284559461805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,16384,0.7036462359958224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,16384,0.19936266210344103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,65536,2.4696044921875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,51200,2.218629413180881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,16384,0.6351386706034342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,12288,0.15015021959940592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,12288,0.5630310906304253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,10240,0.13110310501522487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,12288,0.4921120007832845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,8192,0.29616355895996094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,10240,0.41641510857476133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,51200,2.743096033732096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,8192,0.10618666807810466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,7168,0.2756462097167969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,65536,3.3547689649793835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,7168,0.09382666481865777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,6144,0.225837336646186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,6144,0.08660622437795003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,5120,0.19690577189127603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,5120,0.07308266560236613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,6144,0.22715910275777182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,5120,0.1904062165154351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,3584,0.11582310994466145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,4096,0.05834844377305773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,4096,0.1671288940641615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,3072,0.10972799857457478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,10240,0.40456178453233504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,3584,0.05273422267701891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,2560,0.09864177968766953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,3584,0.13317510816786024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,2048,0.07457955678304036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,3072,0.12086844444274902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,2560,0.04174044397142199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,2560,0.1002417802810669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,1536,0.056863109270731606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,2048,0.036320888333850436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,8192,0.3977946705288357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,1024,0.03952444593111674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,2048,0.08269244432449341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,1536,0.02966044346491496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,768,0.03201777736345927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,1536,0.06611377663082547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,1024,0.02462844385041131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,512,0.025968000292778015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,768,0.022614222433831956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,1024,0.050289776590135365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,256,0.01750666730933719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,768,0.040954666005240545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,512,0.020040889581044514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,128,0.014656000667148165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,256,0.01830755505296919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,512,0.03316088848643833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,64,0.01311377767059538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,128,0.01856355534659492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,32,0.016918222109476726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,128,0.028113779094484117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,3072,0.046479112572140165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,64,0.027973333994547527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,32,0.027837332752015855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,7168,0.2732871108584934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,4096,0.15224711100260416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,65536,0.7393795649210612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,256,0.029143999020258587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,16384,0.4870293405320909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,51200,0.5776284535725912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,16384,0.16634133127000597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,51200,1.5029581917656794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,12288,0.4134533405303955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,12288,0.12794933054182264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,16384,0.6109742058648003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,10240,0.32476621203952366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,65536,2.182538562350803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,10240,0.10868266555998062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,8192,0.2538275453779432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,10240,0.35990222295125324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,8192,0.10018489095899795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,7168,0.24737156762017143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,8192,0.2647368907928467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,51200,2.2694719102647567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,7168,0.07731378078460693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,6144,0.20440267191992867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,7168,0.2414240042368571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,65536,2.7394996219211154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,6144,0.21378489335378012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,5120,0.058719999260372586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,5120,0.18076888720194498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,4096,0.048335111803478666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,3584,0.10878044366836548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,4096,0.12478221787346734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,12288,0.513629330529107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,3584,0.11076622539096409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,3072,0.09582310914993286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,3072,0.03894222113821242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,2560,0.07286666499243842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,6144,0.06875377893447876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,2560,0.03493244449297587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,2048,0.06823466883765326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,2560,0.08254844612545438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,4096,0.1336524486541748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,1536,0.04636889033847385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,2048,0.029457777738571167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,1024,0.035070220629374184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,2048,0.0684115555551317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,768,0.024559999505678814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,1536,0.05465244584613376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,1024,0.021091555555661518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,3584,0.048804445399178394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,512,0.01982133256064521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,768,0.018751111295488145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,1024,0.04174222217665779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,256,0.014307555225160388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,768,0.03446577654944526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,512,0.017072889539930556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,128,0.01235466698805491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,256,0.015840000576443143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,256,0.024146666129430134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,64,0.01036977767944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,128,0.015967110792795818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,32,0.010656889114114972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,128,0.0233262214395735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,5120,0.17331466409895155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,64,0.023394667439990576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,32,0.023243554764323767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,1536,0.025197333759731714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,3072,0.09062933259540135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,65536,0.6407182481553819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,512,0.027387556102540758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,16384,0.4162462287478977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,51200,0.5019786622789171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,16384,0.15001066525777182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,65536,1.5766177707248266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,12288,0.327963564130995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,12288,0.11458399560716416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,16384,0.4974675708346897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,12288,0.37714399231804746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,10240,0.09768533044391209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,8192,0.22101688385009766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,10240,0.2824302249484592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,51200,1.8062195248074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,7168,0.189428448677063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,8192,0.22347556220160592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,7168,0.06932089063856336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,65536,2.3840692308213973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,7168,0.19497599866655138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,5120,0.1274906661775377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,6144,0.06103377872043186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,4096,0.10615289211273193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,51200,1.313645362854004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,5120,0.05618488788604736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,3584,0.09238844447665745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,4096,0.04344444473584493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,10240,0.27792178259955513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,4096,0.11242932743496364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,3584,0.039601776334974505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,3072,0.07679377661810981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,3584,0.10379822386635675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,2560,0.06595644685957167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,3072,0.035585777627097234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,8192,0.07833777533637153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,2048,0.051669332716200084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,3072,0.08575555351045396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,2560,0.0318195554945204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,1536,0.03817066550254822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,6144,0.1710826688342624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,2560,0.07363822062810262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,2048,0.027087999714745417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,1024,0.029151111841201782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,768,0.023208889696333144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,2048,0.06123555368847317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,1536,0.022848889231681824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,512,0.016375111209021676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,1024,0.019158222609096102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,1536,0.04933777782652113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,1024,0.03719377848837111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,6144,0.16643910937839085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,512,0.016195555528004963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,128,0.010053333308961656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,5120,0.13753155867258707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,256,0.014939554863505893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,512,0.024967110819286768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,64,0.008843555218643611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,256,0.02165688905451033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,128,0.01498133275243971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,32,0.009672889278994666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,128,0.021009777983029682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,64,0.020983111527231004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,32,0.020956445071432326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,768,0.017619555195172627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,768,0.030998223357730444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,65536,0.5968720118204752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,65536,1.3099608951144748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,51200,1.0361288918389213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,16384,0.38039112091064453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,16384,0.14639288849300808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,51200,0.4763129022386339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,12288,0.29370755619472927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,16384,0.40522223048739964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,10240,0.23617421256171334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,256,0.012715555727481842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,12288,0.2884142133924696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,65536,2.20576540629069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,8192,0.18088089095221627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,10240,0.10028622547785442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,10240,0.23599643177456328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,7168,0.13015111287434897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,8192,0.07642933395173815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,51200,1.7580773035685222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,8192,0.19149243831634521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,6144,0.11230756176842584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,6144,0.05447644326421949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,5120,0.11565955479939778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,7168,0.1712719996770223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,4096,0.09350755479600693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,5120,0.04842844605445862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,6144,0.1493404441409641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,3584,0.07331733571158515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,4096,0.03959466682540046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,5120,0.12030666404300266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,3072,0.06279022163814969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,3584,0.035226666265063815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,4096,0.09749777449501885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,2560,0.05332000056902567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,12288,0.1074862215254042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,2048,0.04648177822430929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,3072,0.03122133347723219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,3584,0.08796799845165676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,1536,0.03416622347301907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,2560,0.028182221783532038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,3072,0.07649244202507867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,1024,0.023600000474188063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,2048,0.0247662216424942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,2560,0.0659137765566508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,768,0.018391999933454726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,1536,0.02050755586889055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,512,0.01402844488620758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,1024,0.01738133364253574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,1536,0.043506665362252124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,768,0.015838222371207345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,1024,0.033279998434914485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,512,0.014820444915029736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,128,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,7168,0.06307733058929443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,512,0.022312889496485393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,256,0.01352533366945055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,64,0.008211555580298105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,256,0.019582221905390423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,32,0.008277333445019191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,128,0.013591110706329346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,128,0.01864088906182183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,64,0.01882222294807434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,32,0.01874755488501655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,2048,0.05416799916161431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,65536,0.5470248858133951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,256,0.010402667025725046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,768,0.02767555581198798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,65536,1.1176808675130208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,51200,0.8842924435933431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,16384,0.3568133248223199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,51200,0.4239519966973199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,12288,0.24043112330966523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,16384,0.33166578080919057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,12288,0.10444533162646824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,10240,0.19017510943942598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,10240,0.07652444309658475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,65536,1.632631089952257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,10240,0.21117866039276123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,7168,0.13816711637708876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,8192,0.061379558510250516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,8192,0.17626133229997423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,6144,0.12081510490841335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,7168,0.059501330057779946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,5120,0.1015555527475145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,6144,0.04764622118737963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,6144,0.12556889322068956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,4096,0.07292089197370741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,5120,0.0415653321478102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,16384,0.1224035554462009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,3584,0.07042666938569811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,5120,0.1068648894627889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,4096,0.0341635545094808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,4096,0.08394133382373387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,3584,0.030785779158274334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,51200,1.2848089006212022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,3072,0.05470933185683357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,8192,0.1563404401143392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,2560,0.045497778389188975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,3072,0.028764443265067205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,2560,0.025034666061401367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,3072,0.06605777475568983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,2560,0.05651999844445122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,1536,0.02784088916248745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,2048,0.02162577708562215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,7168,0.1434968842400445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,1024,0.020505777663654752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,2048,0.04713066750102573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,1536,0.01811644434928894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,1536,0.03779644436306424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,1024,0.015511110424995422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,1024,0.02878311276435852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,512,0.012997333374288348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,768,0.014369777507252164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,768,0.02439022229777442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,256,0.009308444129096137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,512,0.013452444639470843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,12288,0.25177865558200413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,3584,0.07484533389409383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,128,0.008298666940795051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,512,0.01930488811598884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,256,0.012421333127551608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,64,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,32,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,256,0.017179555363125272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,2048,0.036180443233913846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,128,0.012610666453838348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,128,0.016552888684802584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,64,0.01651822196112739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,32,0.016527111331621807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,768,0.016560888952679105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,65536,0.4151422182718913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,51200,0.3325822353363037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,16384,0.21949599848853218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,51200,0.6946035491095649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,16384,0.10007644361919826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,12288,0.19264533784654406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,65536,1.1349644131130643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,16384,0.2647475666469998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,12288,0.09649866819381714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,10240,0.1563484403822157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,12288,0.1998844411638048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,8192,0.13942221800486246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,51200,0.975444475809733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,7168,0.1205751101175944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,8192,0.053295112318462796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,10240,0.1850586732228597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,65536,1.3560906516181097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,6144,0.08839644326104058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,7168,0.04880977670351664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,8192,0.1861342191696167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,5120,0.07956177658504911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,7168,0.11820356051127116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,6144,0.043803556097878345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,4096,0.08122844166225858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,3584,0.06043644746144613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,6144,0.10434666607115005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,4096,0.02962755494647556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,3072,0.04615111152331034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,5120,0.0875084466404385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,3584,0.027314666244718764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,2560,0.038943999343448214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,3584,0.06334577666388617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,2048,0.03347377644644843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,2560,0.02221511138810052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,3072,0.05522222320238749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,1536,0.02619022296534644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,2560,0.04881599876615736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,2048,0.01958577831586202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,1024,0.016904888881577384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,1536,0.016864000095261466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,10240,0.10434222221374512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,1536,0.032080888748168945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,768,0.013120888835854001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,1024,0.014479110638300577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,512,0.010776888993051318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,1024,0.025182222326596577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,768,0.013351110948456658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,256,0.007923555870850882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,768,0.0207022229830424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,512,0.012416889270146688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,5120,0.03724977705213759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,128,0.006509333435032103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,512,0.016880000631014507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,256,0.011715555356608497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,64,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,32,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,256,0.01482488877243466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,4096,0.07314933008617826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,128,0.011710222396585675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,128,0.014114666316244336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,3072,0.025029333101378545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,64,0.014161777165200977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,32,0.014121777481502958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,2048,0.04035022192531162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,65536,0.39309154616461855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,16384,0.20600799719492593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,51200,0.3048817846510145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,51200,0.7244524425930448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,12288,0.17215822802649605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,65536,0.9893422656589084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,16384,0.09051199754079182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,16384,0.2718053393893772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,10240,0.1554515494240655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,12288,0.07532444265153673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,51200,0.904597282409668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,65536,1.1731013192070854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,10240,0.05983822213278877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,10240,0.15091999371846518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,7168,0.09965955548816258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,8192,0.05030133326848348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,6144,0.08469066354963516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,8192,0.1264302200741238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,5120,0.06997422377268474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,7168,0.11055199967490302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,6144,0.038923554950290255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,6144,0.0947351124551561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,5120,0.03218933277659946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,5120,0.07881333430608113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,3584,0.04807910985416836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,4096,0.026995556222067937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,4096,0.06320799721611871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,3072,0.03963288995954726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,3584,0.024481778343518574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,3584,0.0569653312365214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,2560,0.03179377648565505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,8192,0.12604533301459417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,12288,0.2523075474633111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,3072,0.05011022090911865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,2048,0.02789333462715149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,2560,0.01982666717635261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,7168,0.04132711225085788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,2048,0.017283555534150865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,2560,0.04377333323160807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,1024,0.01640266676743825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,1536,0.01529600057337019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,4096,0.05724000268512302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,768,0.013554666605260638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,1536,0.029382222228580054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,512,0.010159999959998662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,1024,0.01314222233162986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,768,0.012433778080675336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,256,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,768,0.019260444574885897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,512,0.015519110692871941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,128,0.006317333214812809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,256,0.011015999648306104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,256,0.013775111072593264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,64,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,3072,0.022284444835450914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,32,0.005764444255166584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,128,0.010965333216720156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,64,0.013101333545313941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,1536,0.02107288936773936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,32,0.013122667041089801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,2048,0.03639466563860575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,65536,0.5679919984605577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,1024,0.02273599969016181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,65536,0.3118808799319797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,51200,0.44254665904574925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,512,0.011687111523416309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,12288,0.13338400257958308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,16384,0.07930399974187215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,51200,0.2413182258605957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,16384,0.22866222593519422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,10240,0.11630045043097602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,51200,0.7643484539455838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,128,0.01313866674900055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,65536,0.9909413655598959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,8192,0.09803466664420234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,12288,0.1570026609632704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,10240,0.05616977479722765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,7168,0.08302311102549235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,10240,0.13594933350880942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,6144,0.06868000162972344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,8192,0.10844710800382827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,7168,0.03712177938885159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,5120,0.054526223076714404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,7168,0.09470399883058335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,6144,0.03343377841843499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,4096,0.04737866587109036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,5120,0.027886221806208294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,5120,0.07147733370463054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,3584,0.0385519994629754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,16384,0.17777244249979654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,4096,0.02386933399571313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,4096,0.057265778382619224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,3072,0.03443200058407254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,3584,0.02181600034236908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,2560,0.028799109988742407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,3072,0.019839111301634047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,2048,0.023153778579499986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,2560,0.01773155563407474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,3072,0.04472088813781738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,12288,0.06064889166090223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,2560,0.03941422369745042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,1536,0.018235555953449674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,2048,0.01585244470172458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,1024,0.013134222063753339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,8192,0.039741333987977766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,2048,0.03305777907371521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,768,0.01073333372672399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,1536,0.026782222919993933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,1024,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,512,0.008271999657154083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,1024,0.01999288962946998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,768,0.011408888631396823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,768,0.017272889614105225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,256,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,6144,0.08311377631293403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,128,0.005882666756709416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,512,0.014129777749379476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,256,0.012445333103338877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,128,0.010201777848932479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,64,0.005197333378924264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,128,0.011886222494973077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,32,0.005356444252861871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,3584,0.05091555582152473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,64,0.012091555529170565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,32,0.011943110989199745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,1536,0.014151111245155334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,65536,0.5390373335944282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,65536,0.28691111670600045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,51200,0.48955286873711484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,512,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,16384,0.15754577848646376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,256,0.010145777629481422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,51200,0.22904088762071398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,65536,0.8242097960578071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,12288,0.11887199348873562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,16384,0.07086400190989177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,10240,0.0980808867348565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,12288,0.0558462209171719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,8192,0.07704622215694852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,51200,0.6326390902201334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,16384,0.22274221314324272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,10240,0.050423999627431236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,7168,0.06518399715423584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,6144,0.058687998188866504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,12288,0.13923554950290257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,8192,0.03763822383350796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,10240,0.11712711387210423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,5120,0.048560887575149536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,4096,0.0384115543630388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,8192,0.09792800082100762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,6144,0.031457778480317854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,7168,0.08697422345479329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,3584,0.03386311067475213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,5120,0.026743110683229234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,6144,0.07309422228071424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,3072,0.0291244453854031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,5120,0.0617848899629381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,2560,0.02522666586769952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,3584,0.021035555336210463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,4096,0.0498577786816491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,3584,0.0459537771013048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,2048,0.020266667008399963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,2560,0.01759199963675605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,3072,0.03990666733847724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,1536,0.017303110824690927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,2048,0.015503111812803479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,2560,0.03495555453830295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,1024,0.013012444807423485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,2048,0.029574222034878198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,1536,0.013745778136783175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,768,0.011167999770906238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,1024,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,1536,0.023953777220514085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,512,0.010204444328943888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,1024,0.018087999688254464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,768,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,256,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,768,0.015446222490734525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,512,0.01035733355416192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,128,0.005418666534953647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,512,0.012759110993809171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,7168,0.035102221700880266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,256,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,64,0.004774222357405556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,32,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,128,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,256,0.011298666397730509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,128,0.010679110884666443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,4096,0.023307555251651343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,64,0.010701333483060202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,32,0.010759111079904767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,3072,0.018918222851223417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,65536,0.262717326482137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,51200,0.3708933459387885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,51200,0.20464800463782418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,12288,0.09265244669384426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,16384,0.06604266828960843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,16384,0.1559848917855157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,65536,0.6437155405680338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,12288,0.05846755372153389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,10240,0.08480533626344468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,51200,0.5077688958909776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,12288,0.11894133355882432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,8192,0.06400711006588407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,10240,0.04889600144492256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,7168,0.05531822310553657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,8192,0.033886220720079206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,10240,0.10023378001319037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,6144,0.04725333385997348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,8192,0.08041333489947848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,7168,0.030712889300452337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,5120,0.03843911157713996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,7168,0.07407733466890123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,6144,0.027278222971492346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,4096,0.031423999203575984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,6144,0.06196710798475477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,65536,0.39214489195081925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,3584,0.027423111928833857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,5120,0.05214577913284302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,4096,0.01991911066903008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,4096,0.042698668109046094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,3584,0.01828622155719333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,2560,0.019260444574885897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,3584,0.038500444756613836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,3072,0.016619554824299283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,3072,0.034352001216676496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,2048,0.015994666351212394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,2560,0.01516355574131012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,2048,0.01401511165830824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,1536,0.013242666920026144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,2048,0.02510044475396474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,1024,0.009304000271691216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,1536,0.01237066669596566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,1536,0.020630222227838304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,768,0.007892444729804993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,1024,0.015816888875431485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,5120,0.023606222536828782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,512,0.006189333481921091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,16384,0.1174888875749376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,768,0.010208889014191097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,768,0.013758222262064615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,3072,0.02352622151374817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,256,0.004974222017659081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,128,0.004534222185611725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,512,0.009745777481132084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,256,0.009213333328564962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,64,0.003882666842805015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,512,0.011366222467687396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,32,0.004243555582231945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,256,0.010011555420027839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,128,0.009175111022260454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,128,0.009720889230569204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,2560,0.02972088919745551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,64,0.009544889132181803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,32,0.009654222263230218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,65536,0.358387549718221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,65536,0.1848293277952406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,51200,0.2984702322218153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,16384,0.07909955581029256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,51200,0.150344000922309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,12288,0.07218044333987765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,16384,0.04837155673238966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,1024,0.010701333483060202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,10240,0.05996977620654636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,65536,0.5299493471781412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,12288,0.03820888863669501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,16384,0.14897688229878744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,10240,0.0339680016040802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,12288,0.10047555632061428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,7168,0.040692445304658674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,6144,0.0355057782597012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,10240,0.08467911349402533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,7168,0.023153778579499986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,8192,0.06799999872843425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,5120,0.026829333768950567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,6144,0.021162667208247717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,7168,0.06060889032151964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,4096,0.023375999596383836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,6144,0.05423555440372891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,3584,0.020993777447276648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,5120,0.01847111185391744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,5120,0.044646223386128746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,4096,0.016230222251680162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,3072,0.017609778377744887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,4096,0.037857777542538114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,3584,0.015072888798183866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,2560,0.016158221496476066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,3584,0.03428533342149522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,3072,0.013815110756291283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,2048,0.013578666581047906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,3072,0.02954933378431532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,2560,0.012763555679056378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,1536,0.010642666783597736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,2560,0.025646222962273493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,2048,0.011728888584507836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,1024,0.008303110798199972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,8192,0.0492159989145067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,1536,0.010652444428867765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,2048,0.021720889541837905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,768,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,51200,0.41750134362114805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,1536,0.018159111340840656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,512,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,8192,0.025590222742822435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,1024,0.013911111487282647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,768,0.009341333475377824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,256,0.004519111166397731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,512,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,128,0.0038497779104444715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,768,0.01204533295498954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,64,0.0035182221068276297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,512,0.010080888867378235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,32,0.003593777616818746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,128,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,256,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,128,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,64,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,65536,0.27549778090582955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,32,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,65536,0.1714951064851549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,51200,0.20709689458211264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,16384,0.06356444623735216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,51200,0.13797333505418566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,65536,0.39803467856513125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,12288,0.04841777682304382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,16384,0.04683822062280443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,16384,0.10389600197474162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,10240,0.04028088847796122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,1024,0.009710222482681274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,12288,0.03616888986693488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,8192,0.028245333168241713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,12288,0.07725866635640462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,7168,0.027388445205158655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,10240,0.032278223170174494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,8192,0.024493333366182115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,256,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,10240,0.06520533561706543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,6144,0.02199199961291419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,7168,0.02204711072974735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,8192,0.05340266558859083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,5120,0.018473777506086562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,4096,0.015736889508035447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,7168,0.04707377817895678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,6144,0.01983644399378035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,3584,0.014351111319330005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,5120,0.017670222454600863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,6144,0.041088889042536415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,4096,0.015507555670208402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,3072,0.012776888906955719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,3584,0.014480888843536377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,4096,0.028958221276601154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,2560,0.011137777732478248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,3584,0.026214222113291424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,3072,0.013606222139464485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,2048,0.009415999882751042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,2560,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,3072,0.02357599967055851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,1536,0.007861333588759104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,2048,0.011427555647161273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,2560,0.020473778247833252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,1024,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,2048,0.017493334081437852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,1536,0.01034222212102678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,1024,0.00943022221326828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,512,0.00452888881166776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,768,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,1024,0.011730666789743634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,256,0.003555555724435382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,768,0.010149333212110732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,512,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,128,0.003206222214632564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,512,0.00869599978129069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,256,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,64,0.0028533333291610083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,256,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,128,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,32,0.0031475555151700974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,5120,0.0351475543446011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,128,0.007582222421964009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,64,0.007078222102589077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,65536,0.18818755944569907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,32,0.007249777515729268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,51200,0.16145422723558214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,65536,0.1715537839465671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,16384,0.05327999922964308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,768,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,51200,0.13639822271135119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,65536,0.3488479985131158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,12288,0.04131199916203817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,1536,0.014672888649834527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,16384,0.04593777656555176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,51200,0.27139022615220815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,10240,0.0326204432381524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,16384,0.09092799822489421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,10240,0.03205510973930359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,12288,0.06877955463197497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,7168,0.023072888453801472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,8192,0.023906666371557448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,10240,0.05793244308895535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,8192,0.04732444551255968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,6144,0.02032177812523312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,7168,0.02188355558448368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,5120,0.017864889568752713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,7168,0.04236888885498047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,6144,0.019800888167487252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,4096,0.01496355566713545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,6144,0.03713422351413303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,5120,0.017516444126764934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,3584,0.01309866706530253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,4096,0.015508444772826301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,5120,0.03157511022355821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,3072,0.011388444238238864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,4096,0.026483555634816486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,3584,0.014514666464593677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,51200,0.3445448875427246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,2560,0.010425777898894416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,3072,0.013461333182122974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,3584,0.0237137774626414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,2048,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,3072,0.02132622235351139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,2560,0.012435555458068848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,12288,0.035896887381871544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,2048,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,2560,0.018592889110247295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,1024,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,2048,0.01590488851070404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,1536,0.01036266651418474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,768,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,1536,0.013436444103717804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,1024,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,512,0.004436444491147995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,1024,0.010695111420419482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,768,0.009153777526484596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,256,0.003510222252872255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,768,0.00942844483587477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,512,0.008618666893906063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,128,0.003031111011902491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,512,0.008174222376611497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,256,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,64,0.0028799999919202593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,32,0.002895111011134254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,256,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,128,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,128,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,65536,0.14345955848693848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,64,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,32,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,51200,0.11528444290161133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,65536,0.12578133742014566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,8192,0.025384000606007043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,1536,0.007505777809354994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,65536,0.3007555537753635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,51200,0.10154133372836643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,12288,0.030748443471060857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,10240,0.024754666619830664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,16384,0.03547199898295932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,51200,0.2371048927307129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,16384,0.08027555545171101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,10240,0.02472711106141408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,12288,0.059980445437961154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,7168,0.016955556141005624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,8192,0.017842666970358956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,10240,0.05064622229999966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,6144,0.015058666467666626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,8192,0.041386667225095965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,7168,0.016563556260532804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,7168,0.037271999650531344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,6144,0.015113777584499784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,4096,0.01090400003724628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,6144,0.032672888702816434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,5120,0.01349244432316886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,3584,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,4096,0.012078222301271228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,4096,0.023383999864260357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,3072,0.009714666340086196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,3584,0.011391999820868174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,3584,0.021355556117163763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,2560,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,3072,0.01109777804878023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,12288,0.02699377801683214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,8192,0.019125334090656705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,2048,0.0075493330756823225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,2560,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,1536,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,2048,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,2048,0.014410666293568082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,5120,0.013175110850069257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,1024,0.004979555391603046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,1536,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,1536,0.012074666718641916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,768,0.004604444321658877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,1024,0.008558221989207799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,512,0.0038604442444112566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,768,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,5120,0.027801778581407335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,256,0.00315022220214208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,768,0.008674666285514832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,512,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,16384,0.04101422097947862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,512,0.0073626670572492815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,128,0.0026453334010309647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,256,0.007524444825119442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,256,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,64,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,32,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,128,0.007344889144102733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,3072,0.018926221463415358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,128,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,65536,0.09743999772601658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,2560,0.01679733395576477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,64,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,32,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,51200,0.08023644155926175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,16384,0.03166755702760484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,65536,0.12531022230784097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,51200,0.10053688950008816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,65536,0.2574684354994032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,16384,0.034705777963002525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,10240,0.01867822143766615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,16384,0.06791644626193576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,12288,0.026855111122131348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,1024,0.009721777505344814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,8192,0.013518222504191928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,10240,0.02363377809524536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,12288,0.05187288920084635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,7168,0.012230222423871359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,10240,0.04417599903212654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,8192,0.017628444565667045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,6144,0.0110871113008923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,8192,0.03605688942803277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,7168,0.016456888781653512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,5120,0.009645333720578088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,7168,0.03234222200181749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,6144,0.01495822270711263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,4096,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,3584,0.00702044450574451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,6144,0.02842666705449422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,5120,0.013587555951542325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,4096,0.012147555748621622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,5120,0.024359110328886244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,3584,0.011373333632946014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,4096,0.020579554968410067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,2560,0.005719111197524601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,3584,0.01862844493654039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,3072,0.010784889260927836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,2048,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,3072,0.01692800058258904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,12288,0.02179644505182902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,1536,0.004253333227501975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,2560,0.014945778581831189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,51200,0.20347466733720568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,1024,0.003536000019974179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,2048,0.01312355531586541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,1536,0.00904088881280687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,768,0.0031511110977994073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,1536,0.011019555230935415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,1024,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,512,0.0028782222006056043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,768,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,256,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,768,0.008012444608741337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,512,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,128,0.0021893332401911416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,512,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,256,0.0073315559162033936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,64,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,256,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,32,0.0022071111533376905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,3072,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,65536,0.06582488616307576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,64,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,32,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,2560,0.010332444475756751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,51200,0.05282400051752726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,2048,0.009755555954244401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,65536,0.12450044684939915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,16384,0.019296000401178997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,65536,0.2203191121419271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,12288,0.011931555966536203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,51200,0.10007111231486003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,16384,0.0347999996609158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,51200,0.17326577504475912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,16384,0.05815822548336453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,10240,0.012916444076432122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,12288,0.026965333355797663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,8192,0.013019555144839816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,12288,0.04442400071356031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,10240,0.024588444166713293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,7168,0.010771555205186209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,8192,0.018232888645595975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,10240,0.03741688860787286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,6144,0.009789333575301701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,8192,0.030928889910380047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,128,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,7168,0.01688088807794783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,128,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,5120,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,4096,0.007238222493065729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,7168,0.027607111467255488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,6144,0.015657777587572735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,3584,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,6144,0.02437688906987508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,5120,0.021187555458810594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,3072,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,4096,0.01815022197034624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,3584,0.011559999651379056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,2560,0.0053591111467944244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,3584,0.016511110795868766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,3072,0.011054221954610614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,2048,0.004565333326657613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,3072,0.01499644418557485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,1024,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,2560,0.013411555853154926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,1536,0.0041742221348815495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,2048,0.009868444667922126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,1024,0.0035075553589397003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,2048,0.011718221836619906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,1536,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,768,0.0031893334041039148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,1536,0.00999111102686988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,1024,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,512,0.0028284444577164124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,1024,0.008475555313958062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,768,0.008055110772450766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,256,0.002534222271707323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,768,0.007667555577225155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,512,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,5120,0.01402666668097178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,128,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,512,0.006565333240562015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,4096,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,64,0.002128888956374592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,256,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,32,0.0021191111041439907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,256,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,65536,0.06342489189571805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,128,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,51200,0.0510791109667884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,64,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,16384,0.01441511180665758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,12288,0.01348888874053955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,2560,0.010488889283604093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,10240,0.01184177812602785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,65536,0.2201155556572808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,51200,0.1729715532726712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,8192,0.01055111073785358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,7168,0.009921777579519484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,12288,0.044214222166273326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,6144,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,10240,0.03751911057366265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,5120,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,8192,0.030807110998365615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,4096,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,7168,0.027422222826215956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,3584,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,6144,0.02427022159099579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,3072,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,5120,0.02113955550723606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,2560,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,4096,0.01796711153454251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,128,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,2048,0.00453599997692638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,3584,0.01643200053109063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,1536,0.003859555555714501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,1024,0.0032053333189752367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,3072,0.014821334017647637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,768,0.003100444459252887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,2560,0.013415111435784234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,512,0.002623999905255106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,256,0.002469333302643564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,2048,0.011717333561844297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,128,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,1536,0.00998755544424057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,64,0.0021226666867733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,768,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,512,0.006540444576077991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,16384,0.057948443624708385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,32,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,256,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,65536,0.06095288859473335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,51200,0.05059022373623318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,128,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,64,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,12288,0.010434666441546546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,32,0.0052888890107472735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,10240,0.009342222577995723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,65536,0.2208640045589871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,16384,0.05806044737497965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,51200,0.17287645075056288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,8192,0.008647110727098253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,7168,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,12288,0.04433866673045688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,6144,0.0075431110130416016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,5120,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,8192,0.0306915541489919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,4096,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,32,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,5120,0.021194666624069214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,3584,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,1024,0.008316444853941599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,3072,0.007143999967310164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,3584,0.016538666354285348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,2560,0.006445333361625671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,16384,0.013097777962684631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,2048,0.004507555729813046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,3072,0.014888889259762235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,1536,0.003859555555714501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,2560,0.013461333182122974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,1024,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,2048,0.011640888949235281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,768,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,1536,0.010029333333174387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,512,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,1024,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,256,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,7168,0.027395556370417278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,6144,0.024282667371961806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,128,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,512,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,64,0.0018773333479960759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,32,0.00198399999903308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,256,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,4096,0.018057778477668762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,128,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,64,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,32,0.005035555611054103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,10240,0.03737688726849026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,768,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,16384,1.1304115719265408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,16384,3.5788247850206165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,16384,4.102288140190972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,51200,4.17338138156467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,12288,2.9285049438476562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,16384,3.428365283542209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,12288,2.815814124213325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,12288,0.8458791308932834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,10240,0.6886089113023547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,10240,2.4239955478244357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,10240,2.265549341837565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,12288,2.6127840677897134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,8192,0.5571866565280491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,10240,2.146637386745877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,8192,1.9606783125135634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,8192,1.8019235399034288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,7168,0.503369755215115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,8192,1.712180455525716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,51200,11.5325198703342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,7168,1.7039395438300238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,7168,1.5489511489868164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,6144,0.43570934401618105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,7168,1.4614808824327257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,6144,1.65538756052653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,6144,1.2978542115953233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,5120,1.1755457984076607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,5120,0.38871023390028214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,6144,1.2166631486680772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,5120,1.1322790781656902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,4096,0.855188422732883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,4096,0.3146115673912896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,4096,0.8943475087483724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,5120,1.0597830878363717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,3584,0.7740702099270291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,3584,0.27760177188449436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,3584,0.7734097904629178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,4096,0.8149902025858561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,3072,0.760990195804172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,3072,0.2501271035936144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,3584,0.6990719901190864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,3072,0.6787448989020454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,2560,0.6284231079949273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,2560,0.2122622198528714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,3072,0.6204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,2048,0.4227946599324544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,2560,0.5767084227667915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,2048,0.18433333767784965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,1536,0.3415991200341119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,2560,0.521483580271403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,2048,0.5048089027404785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,1536,0.36418133311801487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,1024,0.25576888190375435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,1536,0.14918755160437688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,51200,16.84475368923611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,51200,11.507359822591146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,1024,0.12944000297122532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,768,0.1855600012673272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,1536,0.3362390995025635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,512,0.1357697778277927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,768,0.10455466641320123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,768,0.2321715619828966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,1024,0.2528168890211317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,256,0.1050880021519131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,512,0.18598222732543945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,512,0.0942773355378045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,768,0.23729154798719618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,256,0.08076266447703044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,256,0.16514489385816786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,512,0.1666195525063409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,128,0.07789333661397298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,1024,0.2782222165001763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,64,0.06906400124231975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,2048,0.42934311760796445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,256,0.1324311097462972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,32,0.0690808892250061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,128,0.10048800044589573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,64,0.1585022211074829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,32,0.1605715619193183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,128,0.07770578066507976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,128,0.15851555930243597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,65536,4.134440104166667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,51200,3.2786763509114585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,51200,8.593543158637152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,16384,3.2104070451524525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,16384,3.343245400322808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,16384,0.840551111433241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,65536,10.898968166775175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,12288,2.016015158759223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,12288,0.6775849130418566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,16384,2.7204310099283853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,12288,2.3815280066596136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,51200,12.946801079644096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,51200,9.054549323187933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,12288,1.9588097466362848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,10240,0.5392986403571235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,10240,1.8765636020236547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,10240,1.8175235324435766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,8192,0.44635735617743594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,8192,1.5572355058458116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,10240,1.6491253111097548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,8192,1.439807997809516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,65536,15.989376491970487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,7168,1.2172782686021593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,7168,0.3858453432718913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,8192,1.3266639709472656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,65536,11.406823900010851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,6144,0.9973946677313911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,6144,0.3475200070275201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,6144,1.0914817386203342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,5120,0.8348204294840494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,5120,0.3048524326748318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,7168,1.1504631042480469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,5120,0.900728013780382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,4096,0.6860408782958984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,6144,0.980929798550076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,5120,0.7771769099765353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,4096,0.709596421983507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,4096,0.24607732560899523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,3584,0.5450106726752387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,7168,1.3210586971706815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,4096,0.6451724370320638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,3584,0.6415893236796061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,3072,0.47964800728691953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,2560,0.45533153745863175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,3072,0.5183084275987413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,3584,0.5502284367879232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,2560,0.4469920264350043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,2048,0.3332666556040446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,3072,0.47928799523247617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,2560,0.1930453379948934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,2560,0.41568888558281797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,2048,0.36798755327860516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,1536,0.24659556812710234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,2048,0.1441884438196818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,3584,0.22079643938276503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,1024,0.18443022833930123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,1536,0.11583467324574788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,3072,0.1952568954891629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,1536,0.37983200285169816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,768,0.1376968887117174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,1024,0.2252577675713433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,2048,0.3451226552327474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,512,0.10056977801852757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,768,0.08584622542063396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,1024,0.19879999425676134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,768,0.1698862181769477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,1536,0.2675475544399685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,256,0.07519111368391249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,512,0.07532266775767009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,768,0.1674177779091729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,128,0.06318666537602742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,512,0.1458968851301405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,256,0.13070489300621882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,256,0.06525333060158624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,512,0.13384800487094456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,128,0.06333955791261461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,256,0.10704000128640069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,32,0.053216000398000084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,128,0.12450933456420898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,128,0.08129955662621392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,64,0.12525333298577204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,32,0.1283200052049425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,1024,0.10289777649773492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,64,0.05094844433996412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,65536,1.2987581888834636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,51200,0.9824711481730143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,51200,2.6922488742404513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,16384,1.0710568957858615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,65536,3.456765280829536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,16384,0.28232622146606445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,16384,1.1267076068454318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,51200,4.026069217258029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,12288,0.7102710935804578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,16384,0.8338861995273166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,51200,2.9457581837972007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,12288,0.22648178206549752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,12288,0.7113306787278918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,10240,0.54875734117296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,10240,0.18644355403052437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,12288,0.6202898025512695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,65536,3.8120295206705728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,8192,0.47685956954956055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,10240,0.6493502193027073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,65536,5.9845021565755205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,10240,0.5190426508585612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,8192,0.1485217809677124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,8192,0.4605706532796224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,7168,0.1337288882997301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,6144,0.29912445280287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,8192,0.42316799693637425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,7168,0.4070417881011963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,6144,0.11723466714223225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,5120,0.24573334058125815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,7168,0.3656746546427409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,4096,0.20307555463578966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,5120,0.10120089186562432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,6144,0.3119857841067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,5120,0.293832884894477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,4096,0.08454400300979614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,3584,0.19472622871398926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,5120,0.26215733422173393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,4096,0.24327733781602648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,3584,0.07519555754131742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,3072,0.15358044041527644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,7168,0.4023244380950928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,3584,0.2036879989835951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,4096,0.21851644251081678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,2560,0.12448978424072266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,3072,0.06894310977723864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,3072,0.1767902241812812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,3584,0.1908728943930732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,2048,0.11661156018575032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,6144,0.35367290178934735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,2560,0.05973333120346069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,2560,0.14909422397613525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,1536,0.0829004446665446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,3072,0.16829333040449354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,1024,0.06408266888724433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,2048,0.1237306727303399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,2048,0.05197244551446703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,1536,0.042372445265452065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,1536,0.09815377659267849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,2048,0.12153511577182347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,1024,0.036397334602144026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,1536,0.09772888819376628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,1024,0.07518577575683594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,512,0.03668444355328878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,768,0.032877332634396024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,768,0.06057066387600369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,1024,0.07568799787097506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,256,0.02475111186504364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,512,0.02998044424586826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,512,0.051445331838395864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,128,0.021874666213989258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,256,0.027099554737408955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,256,0.04642044504483541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,512,0.053220444255405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,64,0.02350577712059021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,128,0.026424888107511733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,128,0.04398933384153578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,256,0.04191289014286465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,32,0.024930665890375774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,2560,0.14316977394951716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,128,0.03219733304447598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,64,0.04418666826354133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,32,0.0447599987188975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,768,0.04677955640686882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,768,0.06468977530797322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,65536,1.0263030793931749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,51200,0.7869626681009928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,51200,2.139320797390408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,16384,0.8822782304551867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,65536,2.737209743923611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,16384,0.2365644507937961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,16384,0.7762248780992297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,12288,0.49192534552680117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,51200,3.304973390367296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,16384,0.6510675748189291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,51200,2.3099937438964844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,12288,0.17019732793172201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,12288,0.5615244441562229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,12288,0.4675360255771213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,10240,0.4768222173055013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,10240,0.14549867312113443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,8192,0.3834515677558051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,65536,4.408923678927951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,65536,2.9496977064344616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,8192,0.11882489257388645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,8192,0.3579022354549832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,10240,0.39632709821065265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,7168,0.35041332244873047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,7168,0.10583555698394775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,6144,0.2641226715511746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,7168,0.3235422240363227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,7168,0.2978293365902371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,6144,0.09071289168463813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,6144,0.28421245680914986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,5120,0.2347644435034858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,6144,0.24343289269341362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,4096,0.17623821894327799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,5120,0.07823555337058173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,10240,0.4571297963460286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,5120,0.2177520063188341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,4096,0.06521777974234687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,3584,0.16165244579315186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,5120,0.20490311251746285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,3584,0.05970578061209785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,3072,0.11531289418538411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,4096,0.18473333782619902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,3584,0.1538782252205743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,8192,0.32653334405687123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,2560,0.09831200043360393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,3072,0.054120000865724355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,3072,0.13390666908688015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,2048,0.0929795569843716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,2560,0.0468355549706353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,2560,0.11516711446974014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,2560,0.11310577392578125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,2048,0.04018933243221707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,2048,0.09547377957238092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,1536,0.07108977768156263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,2048,0.09616711404588486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,1536,0.03353777858946059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,4096,0.1749591165118747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,1024,0.04936710993448893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,1536,0.07570222351286146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,768,0.040435555908415056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,1024,0.028247998820410833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,1536,0.07828355497784086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,1024,0.058343999915652804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,512,0.030794666873084173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,3584,0.1493404441409641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,768,0.046428445312711925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,768,0.025851555996470984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,1024,0.06217600239647759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,256,0.02623111175166236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,512,0.023592000206311543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,512,0.03881066706445482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,128,0.015184000134468079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,256,0.020815999971495736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,768,0.05112533436881172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,64,0.020863110820452373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,256,0.035049776236216225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,512,0.04354844490687052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,32,0.021464889248212177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,128,0.02051999999417199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,128,0.03275200062327915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,256,0.034345779154035784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,64,0.03296444482273526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,128,0.026203556193245783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,3072,0.13274933232201472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,65536,0.7893582450018989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,32,0.03278399838341607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,51200,1.6912444432576497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,65536,2.2699627346462674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,16384,0.5782924228244358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,51200,0.6347235573662652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,16384,0.18722221586439344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,12288,0.5357226795620388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,16384,0.6299635569254557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,51200,2.7759005228678384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,12288,0.1449271175596449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,16384,0.5288862122429742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,12288,0.4875209066602919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,10240,0.3422871165805393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,65536,3.486273871527778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,10240,0.12337333626217312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,65536,2.3937369452582464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,12288,0.39519556363423664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,8192,0.28793332311842174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,10240,0.37019022305806476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,51200,1.8782800038655598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,8192,0.10811110999849106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,7168,0.2561368942260742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,8192,0.3272248903910319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,10240,0.33096088303460014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,7168,0.09825866752200657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,6144,0.22249333063761392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,6144,0.07748888598548041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,7168,0.32920532756381565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,8192,0.2686799897087945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,5120,0.1798177825080024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,7168,0.2351155545976427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,4096,0.14338133070203993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,5120,0.06730400191413032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,5120,0.20007288455963135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,3584,0.10409688949584961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,4096,0.05778666337331136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,4096,0.1724639998541938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,5120,0.17688443925645617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,3584,0.05079466766781277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,4096,0.14216089248657227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,3584,0.12852622403038874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,3072,0.11630845069885254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,2560,0.0846524437268575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,3072,0.04472622275352478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,3072,0.11232177416483562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,6144,0.30951200591193306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,2560,0.0399253335263994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,2560,0.09838489029142593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,2048,0.07070310910542806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,6144,0.20246399773491752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,1536,0.05330222182803684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,2048,0.03458844290839301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,2560,0.09730399979485406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,2048,0.08173155784606934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,1024,0.03564711080657111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,3072,0.10496177938249375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,1536,0.028677334388097126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,1536,0.0644364423222012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,2048,0.08310755756166247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,768,0.027965333726671007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,1024,0.023880889018376667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,1536,0.06823199987411499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,1024,0.049535112248526685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,512,0.02088711162408193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,768,0.02164177762137519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,1024,0.05323644479115804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,256,0.015728889240158927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,512,0.019730667273203533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,768,0.04494844542609321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,512,0.03258400162061056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,128,0.014516444669829475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,256,0.017882666654056974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,512,0.03791111045413547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,256,0.028943111499150593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,64,0.012454222473833295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,128,0.027507555153634813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,128,0.01870577699608273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,32,0.012179555164443122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,256,0.030288888348473444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,3584,0.1326026651594374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,64,0.027870221270455256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,128,0.023361777265866596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,32,0.027699554959932964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,768,0.04015288750330607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,65536,0.6574968761867946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,51200,0.5300524499681261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,16384,0.6466675334506565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,51200,1.7023742463853624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,65536,1.9715156555175781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,16384,0.15313511424594456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,16384,0.7326035499572754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,12288,0.3968177901373969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,51200,2.125724368625217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,51200,1.5713297526041667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,10240,0.29166311687893337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,12288,0.11986045042673747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,65536,2.769887076483833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,16384,0.4786391258239746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,12288,0.3750844531589084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,65536,2.024486117892795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,10240,0.10281778044170803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,8192,0.2388542228274875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,12288,0.33548887570699054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,8192,0.09120088815689087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,7168,0.22193333837721083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,8192,0.2501893308427599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,10240,0.3294160101148817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,7168,0.075436446401808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,8192,0.2244497670067681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,7168,0.2271866798400879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,5120,0.14491289191775852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,7168,0.2001111110051473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,6144,0.18470222420162627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,4096,0.11365866661071777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,5120,0.056623107857174344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,6144,0.17374844021267363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,5120,0.1523724397023519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,4096,0.04642311069700453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,4096,0.12141511175367568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,3584,0.09879644711812337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,10240,0.3810719913906521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,4096,0.1229742235607571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,3072,0.08697511090172662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,3584,0.042062222957611084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,3584,0.10725777679019505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,6144,0.1731599966684977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,2560,0.06798222329881456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,3072,0.03748266564475165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,3584,0.10883377657996283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,3072,0.09433422485987346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,6144,0.06769511434766981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,2048,0.05864266554514567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,2560,0.03351022137535943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,2560,0.0814337796635098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,1536,0.042931556701660156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,3072,0.09774666362338597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,2048,0.02857066525353326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,5120,0.1601297722922431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,2048,0.06746577554278903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,2560,0.0832186672422621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,1024,0.03168088859981961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,1536,0.02402399977048238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,1536,0.05349422163433499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,2048,0.07207733392715454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,768,0.024704888463020325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,1536,0.05877600113550822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,512,0.018059555027219985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,768,0.01850044396188524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,768,0.03362577822473314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,1024,0.045660446087519325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,256,0.013094222380055321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,768,0.039697776238123574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,512,0.016765332884258695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,512,0.027220444546805486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,128,0.013774221969975365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,256,0.015407111909654407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,256,0.024136000209384497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,512,0.03337333268589444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,64,0.012245333029164208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,128,0.015483554866578845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,32,0.013095999757448832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,256,0.02595377796226078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,64,0.023019555542204116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,32,0.022878222995334204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,1024,0.04139022363556756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,128,0.02288088864750332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,65536,0.6333884663052028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,128,0.0196142229768965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,65536,1.5891475677490234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,1024,0.020940444535679285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,16384,0.4052071041531033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,51200,0.49538400438096786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,16384,0.43685600492689347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,65536,2.300936804877387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,16384,0.15205955505371094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,51200,1.3729172812567816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,12288,0.3070720036824544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,12288,0.1272533337275187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,16384,0.41528889867994523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,65536,1.8418106502956813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,10240,0.27291732364230686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,12288,0.32389865981207955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,10240,0.10067644384172227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,12288,0.3062888781229655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,51200,1.9002816942003038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,51200,1.4212506612141926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,10240,0.33603376812405056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,8192,0.2505999935997857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,8192,0.07481777667999268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,10240,0.25571733050876194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,7168,0.2189404434627957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,6144,0.18671822547912598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,7168,0.06658755408393012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,8192,0.212263994746738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,7168,0.22179555892944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,8192,0.20886311266157362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,6144,0.05889866749445597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,6144,0.16186132695939806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,5120,0.14930310514238146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,7168,0.1835146612591214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,5120,0.050254222419526845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,4096,0.11624977323744033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,3584,0.0986506674024794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,4096,0.04163288738992479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,4096,0.10960711373223199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,5120,0.13611466354793972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,5120,0.13703821765051946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,3584,0.037745777103636004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,3072,0.08254133330451117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,3584,0.09976266490088569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,2560,0.06846488846672906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,3072,0.03375555409325494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,3584,0.09943288564682007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,2048,0.05867289172278511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,2560,0.0732862220870124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,4096,0.11237866348690456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,3072,0.08806488911310832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,3072,0.08546488814883763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,1536,0.04199111130502489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,2048,0.026084444589085046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,2560,0.07735733191172282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,2048,0.06013422542148166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,1024,0.027481777800454035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,1536,0.02164799968401591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,2048,0.06439555353588529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,1536,0.048491554127799145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,768,0.022151110900772944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,6144,0.1634062263700697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,1024,0.01814399990770552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,1024,0.0370124446021186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,512,0.01681422193845113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,1536,0.05334044496218363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,1024,0.04228355487187704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,768,0.03077510992685954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,256,0.01291911138428582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,512,0.015321777926550971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,512,0.02453688946035173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,768,0.035748445325427584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,128,0.010011555420027839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,256,0.01423288881778717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,256,0.02130222154988183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,512,0.03088444471359253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,2560,0.030415111117892798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,64,0.00871288859181934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,32,0.008968888885445064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,256,0.023864888482623633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,128,0.014135999812020196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,128,0.020849777592553034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,64,0.020959999826219346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,32,0.020599111914634705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,128,0.018277333842383493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,768,0.016591999265882704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,65536,0.5304897626241049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,65536,1.2633173200819228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,51200,1.03365601433648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,16384,0.3488231235080295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,51200,0.432894229888916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,51200,1.4295786751641169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,16384,0.40196532673305935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,12288,0.3071315553453234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,16384,0.13071289327409533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,65536,1.9678950839572484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,16384,0.36509066157870823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,10240,0.24517244762844512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,12288,0.2853662172953288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,65536,1.6020133760240343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,10240,0.08778933021757339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,8192,0.16237333085801867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,12288,0.27509156862894696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,8192,0.0658551123407152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,8192,0.19552444087134468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,10240,0.2335075537363688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,7168,0.16773600048489043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,6144,0.1444044510523478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,7168,0.058740443653530545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,8192,0.191666669315762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,7168,0.16142133871714273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,6144,0.051048888100518115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,7168,0.16723555988735625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,5120,0.11571555667453343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,6144,0.16797867086198595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,12288,0.11313154962327744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,4096,0.09293511178758408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,6144,0.14342310693528917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,5120,0.11961778004964192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,51200,1.2390773561265733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,3584,0.08298399713304308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,4096,0.09620800283220078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,5120,0.12174399693806966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,3584,0.034027556578318276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,3072,0.06910489002863567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,3584,0.08443555566999648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,4096,0.10565777619679768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,2560,0.056976000467936196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,3072,0.030435555511050757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,3584,0.09141155746248032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,3072,0.0747706691424052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,2048,0.04775911238458422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,2560,0.02754488918516371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,2560,0.0641804469956292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,3072,0.0801431139310201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,1536,0.03455822335349189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,2048,0.02404977712366316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,2048,0.053723557127846606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,5120,0.043706665436426796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,1024,0.02510844502184126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,10240,0.2476844522688124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,1536,0.042600890000661217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,1536,0.020212444994184706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,2048,0.06028622388839722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,4096,0.036545776658587985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,768,0.02036888897418976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,1024,0.017417778571446735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,1536,0.04924355612860786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,1024,0.03271377748913235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,768,0.015840889679061044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,768,0.02738133403990004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,256,0.010647111468844943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,512,0.014373333917723762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,1024,0.039451556073294744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,768,0.03290133343802558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,512,0.02200177808602651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,128,0.00977955593003167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,256,0.013376889129479727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,256,0.019163555569118924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,512,0.027036445008383855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,64,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,32,0.008604444563388824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,128,0.013800000150998434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,128,0.018372444642914664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,256,0.021741333934995864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,64,0.018564444449212816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,2560,0.06993955373764038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,32,0.018280888597170513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,512,0.015266666809717814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,65536,0.46643378999498153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,65536,1.2148577372233074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,51200,0.9745101928710938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,51200,0.3579271104600694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,128,0.01726222203837501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,16384,0.33862490124172634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,51200,1.27473693423801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,65536,1.750071949428982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,65536,1.32624446021186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,16384,0.1074826651149326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,12288,0.25209776560465497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,16384,0.3621991210513645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,51200,0.9485538270738391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,12288,0.08272977670033772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,10240,0.16687822341918945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,16384,0.29541156027052135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,12288,0.2571457756890191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,8192,0.12682399484846327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,10240,0.07060888740751478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,10240,0.1996391084459093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,12288,0.23519465658399794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,8192,0.061512000030941434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,8192,0.15924533208211264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,10240,0.20742311742570665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,6144,0.10732711023754543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,7168,0.05242400036917793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,8192,0.15696177217695448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,7168,0.14151022169325086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,5120,0.08919466866387261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,6144,0.044377777311537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,6144,0.12377777364518906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,7168,0.13839111063215467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,4096,0.07168355253007677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,5120,0.03867111272282071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,5120,0.10139999787012736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,6144,0.11871555116441514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,3584,0.05587644709481133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,4096,0.03257955445183648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,4096,0.08283555507659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,5120,0.10173333353466457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,3072,0.053879112005233765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,3584,0.028879112667507593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,3584,0.0726720028453403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,4096,0.08397510978910659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,2560,0.03823022378815545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,7168,0.13419911596510145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,3072,0.02622933354642656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,3072,0.0640853312280443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,3584,0.0758897794617547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,2048,0.033960888783137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,2560,0.023941333095232647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,3072,0.06829155815972222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,1536,0.027851555082533095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,2560,0.05520533190833198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,2048,0.020670221911536325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,1024,0.018999111321237352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,2048,0.046345776981777616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,2560,0.05863200293646919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,1536,0.01773866679933336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,768,0.015507555670208402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,2048,0.050498667690489024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,1024,0.015396444333924187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,1536,0.042416890462239586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,512,0.011388444238238864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,768,0.02386666668785943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,768,0.013918222652541267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,256,0.008665777742862701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,512,0.012792000340090858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,512,0.019115555617544387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,768,0.027506666051016912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,128,0.007416889071464539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,256,0.016915554801623028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,256,0.011997333003415002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,128,0.016172443826993305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,128,0.012051555845472546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,256,0.020928000410397846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,64,0.006572444405820634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,32,0.006830222076839871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,128,0.01610577768749661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,64,0.016193777322769165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,32,0.016179554992251925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,1536,0.03753688931465149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,1024,0.028228445185555354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,1024,0.03310044606526693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,65536,0.40776978598700625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,65536,0.9935323927137586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,512,0.023269333773189124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,51200,0.8520097732543945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,16384,0.2657724486456977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,51200,0.3266826735602485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,16384,0.2619697782728407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,16384,0.10066133075290257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,65536,1.3623893525865343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,12288,0.21138221687740752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,51200,0.9899529351128472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,65536,1.1005662282307942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,16384,0.27635465727912056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,10240,0.17624266942342123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,12288,0.07887999878989325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,51200,0.8647857771979438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,12288,0.19569067160288492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,8192,0.14258489343855116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,10240,0.06535644663704766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,10240,0.17060089111328125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,8192,0.05110222101211548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,12288,0.20055733786688912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,7168,0.1235777801937527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,8192,0.14110578431023493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,6144,0.10081688563028972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,7168,0.04409866531689962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,10240,0.16990666919284395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,8192,0.14791556199391684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,7168,0.11698844697740342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,5120,0.0833244456185235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,6144,0.03803911142879062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,6144,0.10277689165539211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,7168,0.12092088328467475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,4096,0.05766666597790188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,5120,0.08553155263264973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,6144,0.10740800036324395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,3584,0.04825777808825175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,4096,0.028604444530275133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,5120,0.09038222498363918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,4096,0.06911110877990723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,3072,0.0448968874083625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,3584,0.02479377720091078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,3584,0.06308711237377591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,4096,0.07579200135336982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,3072,0.022651554809676275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,3072,0.054437332683139376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,3584,0.06753155257966784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,2048,0.031641778018739485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,2560,0.02055644492308299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,2560,0.04745955599678887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,2048,0.01773244473669264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,1536,0.022943110929595098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,2048,0.03972711165746053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,5120,0.03287555442916022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,1024,0.01755288905567593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,2048,0.04450311263402303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,1536,0.03199111090766059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,1536,0.03717422154214647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,1024,0.013541333377361298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,1024,0.024323556158277724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,768,0.013352889153692456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,1024,0.029395557112163965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,768,0.01260088880856832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,512,0.010737777584128909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,768,0.020657777786254883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,3072,0.06038577689064873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,768,0.024845333562956914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,256,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,512,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,512,0.01658933361371358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,128,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,256,0.011018666956159802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,512,0.021073778470357258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,2560,0.051664001411861844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,64,0.00608711110221015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,256,0.014708444476127625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,32,0.0063075555695427795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,128,0.01408266690042284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,128,0.01109688894616233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,1536,0.015602666470739575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,256,0.019408888287014432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,64,0.014093332820468478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,32,0.014039999908871122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,128,0.014747555057207743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,2560,0.03489777776930068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,65536,0.3500799867841933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,65536,0.8061164220174154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,16384,0.23316711849636504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,51200,0.6335777706570095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,51200,0.27788710594177246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,16384,0.0931697752740648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,65536,1.1304133733113606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,12288,0.17983022001054552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,51200,0.8535751236809624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,16384,0.2438275549146864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,10240,0.13884800010257298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,12288,0.07190222210354276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,65536,1.0558044645521376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,51200,0.8795919948154025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,16384,0.2733439869350857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,12288,0.18331733014848497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,8192,0.10767289002736409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,10240,0.05945599741405911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,10240,0.15362311734093562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,7168,0.09701244698630439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,8192,0.04524622360865275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,12288,0.20300532711876762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,6144,0.08535289102130467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,8192,0.13153778182135686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,7168,0.044581333796183266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,10240,0.1991973320643107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,7168,0.10752089156044854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,8192,0.13451911343468562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,6144,0.03865155577659607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,6144,0.09054933653937446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,4096,0.05204622281922234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,7168,0.11784799893697102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,5120,0.031081779135598078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,5120,0.08056088950898913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,6144,0.10337066650390625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,4096,0.02608000073168013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,3584,0.04507022102673849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,4096,0.062337776025136314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,5120,0.0885573360655043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,3072,0.03835555579927232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,3584,0.024102222588327196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,4096,0.07464710871378581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,3584,0.055926223595937095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,2560,0.0322106679280599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,3072,0.021856000026067097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,3072,0.04959288901752896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,3584,0.06580266687605116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,2048,0.02516533268822564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,2560,0.019900444481107924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,2560,0.0428071121374766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,3072,0.05858933263354831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,1536,0.02108088963561588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,2048,0.01723377737734053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,2048,0.03585955500602722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,5120,0.06927200158437093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,1024,0.014704000618722705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,1536,0.014831999937693277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,1536,0.028696000576019287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,2048,0.043990221288469106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,768,0.011605333122942181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,1024,0.013088888592190213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,1536,0.036122666464911565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,512,0.009631111390060848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,768,0.018872888551818002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,768,0.012439111040698158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,768,0.024028443627887305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,512,0.01166044423977534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,256,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,512,0.020863110820452373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,256,0.013470222552617392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,128,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,256,0.010713777608341642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,256,0.017251556118329365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,2560,0.0513919989267985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,64,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,128,0.01279022213485506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,128,0.010743111371994019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,32,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,1024,0.022447999980714586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,64,0.012780444489585029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,128,0.014373333917723762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,32,0.012895111408498553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,1024,0.02834933333926731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,512,0.015338665909237333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,65536,0.6758835580613879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,65536,0.30591289202372235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,51200,0.551777786678738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,16384,0.1933048831091987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,51200,0.24322666062249076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,12288,0.13883554935455322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,16384,0.07796088854471843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,65536,1.0156986448499892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,16384,0.25358398755391437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,51200,0.7964168654547797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,10240,0.11648444334665935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,12288,0.06049066781997681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,65536,0.8727777798970541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,16384,0.23264710108439127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,51200,0.6915635532803006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,12288,0.16347466574774847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,8192,0.09505955378214519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,10240,0.05106755428844028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,10240,0.14186755816141763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,7168,0.08099022176530626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,12288,0.17253155178493926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,8192,0.03942222065395779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,8192,0.10797688696119521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,6144,0.07015199793709649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,7168,0.037680887513690524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,7168,0.09499733315573798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,5120,0.06132000022464328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,8192,0.11628444989522298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,6144,0.031194667021433514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,6144,0.0830284423298306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,4096,0.04765955607096354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,7168,0.1038844452963935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,5120,0.027460444304678176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,5120,0.0690586699379815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,6144,0.09057866864734226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,4096,0.023583110835817125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,4096,0.056346668137444385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,3072,0.03386044502258301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,3584,0.021368889345063105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,3584,0.050907555553648204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,3072,0.019695111446910434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,3072,0.0446266664399041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,2560,0.02908266584078471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,3072,0.051868445343441434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,10240,0.14465955893198648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,2560,0.038896001047558255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,2048,0.025384000606007043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,2560,0.017688888642523024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,2560,0.045662220981385976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,2048,0.01569333341386583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,1536,0.01962222158908844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,2048,0.03276355399025811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,3584,0.040475553936428495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,2048,0.03921688927544488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,1536,0.013952889376216464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,1024,0.01443822185198466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,5120,0.07752177450391981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,1536,0.02644088864326477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,768,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,4096,0.06518933508131239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,1024,0.012283555335468717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,1536,0.03237866693072849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,1024,0.020024889045291476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,3584,0.05848977963129679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,512,0.008972444468074376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,768,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,256,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,1024,0.025399110383457605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,512,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,768,0.02200800014866723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,512,0.013914666242069669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,128,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,256,0.010255110760529837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,256,0.012277333272827996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,512,0.018592889110247295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,64,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,32,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,128,0.010416888528399997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,128,0.011985777980751462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,256,0.01622222198380364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,64,0.011759999725553723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,128,0.012713777522246042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,32,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,65536,0.2800435490078396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,65536,0.6331049071417915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,51200,0.48179202609592015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,16384,0.17876177363925508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,65536,0.7617422209845649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,51200,0.22176622019873726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,51200,0.6168470912509495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,16384,0.18654844495985243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,65536,0.7287395795186361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,12288,0.12707911597357854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,10240,0.10811377896202935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,51200,0.5752035776774088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,12288,0.059643553362952344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,16384,0.19093422094980875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,12288,0.13803999953799778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,8192,0.08865422010421753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,10240,0.050035556157430015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,10240,0.11875643995073105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,12288,0.14597511291503906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,7168,0.07668888568878174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,8192,0.09286133448282878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,10240,0.12200710508558486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,768,0.01716800034046173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,6144,0.0635511080423991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,7168,0.03346844514211019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,7168,0.0818328857421875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,8192,0.09949866930643718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,5120,0.052573333183924355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,6144,0.02966044346491496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,6144,0.0713004469871521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,7168,0.08842577536900838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,4096,0.03820800118976169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,16384,0.06951289044486152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,5120,0.06010755565431383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,6144,0.07783289088143243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,3584,0.032551109790802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,4096,0.021872000561820135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,5120,0.06621778011322021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,4096,0.048559112681282886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,3072,0.028973334365420874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,3584,0.020230222079488967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,3584,0.043775998883777194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,2560,0.026233777403831482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,3072,0.01862133377128177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,3072,0.038663112454944186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,8192,0.0380675560898251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,2048,0.019865777757432725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,2560,0.017126222451527912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,2560,0.03369422091378106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,2048,0.015281778242852954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,2048,0.027895109521018133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,2560,0.03916533456908332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,1536,0.015824000040690105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,1536,0.01278222186697854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,5120,0.02626222206486596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,2048,0.03330844309594896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,1024,0.0123511114054256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,1536,0.022843556271659002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,768,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,1024,0.011108444796668159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,1536,0.028004444307751123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,1024,0.017863111363516915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,512,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,4096,0.05633688635296292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,768,0.015141333142916361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,1024,0.02221511138810052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,3584,0.050739556550979614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,256,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,512,0.010088000032636855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,512,0.012453333371215396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,768,0.019582221905390423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,128,0.004859555512666702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,3072,0.045068442821502686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,256,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,512,0.01829600003030565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,256,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,64,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,32,0.004588444613748127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,128,0.009721777505344814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,128,0.01071466671095954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,256,0.015172445111804538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,64,0.01038933296998342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,128,0.012074666718641916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,32,0.01074222226937612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,65536,0.4502711296081543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,65536,0.2388711240556505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,51200,0.38663199212816024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,16384,0.13409777482350668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,65536,0.6385146776835123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,51200,0.5004915661282009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,16384,0.06146577994028727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,65536,0.7046142154269748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,16384,0.16442222065395778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,12288,0.10018933481640285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,51200,0.5567110909356011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,16384,0.18381422095828584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,12288,0.04725689027044508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,10240,0.0810613301065233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,12288,0.12072177728017171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,8192,0.06727821959389581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,10240,0.042020443412992686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,12288,0.14545155896080866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,10240,0.09835200177298652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,768,0.010703110860453712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,7168,0.0548000004556444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,8192,0.03152622116936578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,8192,0.0806186662779914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,10240,0.11877866586049397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,6144,0.0503057771258884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,51200,0.19383645057678223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,7168,0.028551111618677776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,7168,0.07060799996058147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,5120,0.039992888768514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,8192,0.09719555245505439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,6144,0.025855110751258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,4096,0.03269688950644599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,6144,0.06158666478263008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,7168,0.0863493349817064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,3584,0.028526223368114893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,5120,0.051382223765055336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,6144,0.07567555374569364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,4096,0.01937866707642873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,4096,0.042429334587521024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,3072,0.024126221736272175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,3584,0.01770666738351186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,3584,0.03859022259712219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,4096,0.054639998409483165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,2560,0.020480889413091872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,3072,0.016363556186358135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,3072,0.03379022081693014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,3584,0.04933244321081373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,2048,0.01738133364253574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,2560,0.015005333556069268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,2560,0.029743999242782593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,3072,0.04370133413208855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,1536,0.013173333472675748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,2048,0.013761777844693927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,2048,0.024862221545643274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,2560,0.03820977939499749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,1024,0.009481777747472128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,1536,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,2048,0.03235377868016561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,768,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,5120,0.02254044512907664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,1536,0.027287110686302185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,1024,0.015744000673294067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,512,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,1024,0.021969777014520433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,768,0.010325333310498131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,5120,0.06569244464238484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,256,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,512,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,512,0.011360888679822287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,768,0.019588443968031142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,128,0.00424888895617591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,256,0.009078222016493479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,256,0.009923555784755284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,512,0.016539555456903245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,64,0.004056888735956616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,128,0.009143999881214565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,128,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,32,0.00423911131090588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,256,0.015018666783968607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,64,0.009379555781682333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,128,0.011352000137170156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,1536,0.020495999190542433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,32,0.009532444179058075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,65536,0.3262142340342204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,1024,0.010728889041476779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,65536,0.1831937763426039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,768,0.013542222479979197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,51200,0.30099911159939235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,16384,0.0979457762506273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,65536,0.5291848712497288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,51200,0.14572532971700033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,51200,0.4151804447174072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,16384,0.048487110270394214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,12288,0.0714320010609097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,16384,0.1621697743733724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,10240,0.0605600012673272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,65536,0.5610319773356119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,12288,0.03802311089303758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,51200,0.44347643852233887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,16384,0.14757333861456975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,12288,0.09973422023985122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,8192,0.04965244399176704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,10240,0.03351199958059523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,10240,0.08760977453655666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,7168,0.04342577854792277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,12288,0.11584799819522434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,8192,0.026541334059503343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,6144,0.037301331758499146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,8192,0.06753689050674438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,10240,0.1034168865945604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,7168,0.02313511073589325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,5120,0.030773331721623737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,7168,0.060601777500576444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,8192,0.08043022288216485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,6144,0.02106399999724494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,4096,0.02502222193611993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,6144,0.05306666758325365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,5120,0.01822933389080895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,7168,0.07094133562511869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,3584,0.022491556074884202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,5120,0.044509334696663745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,6144,0.06168533033794827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,4096,0.016018667154841952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,3072,0.019522666931152344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,4096,0.037061333656311035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,5120,0.05418222149213155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,3584,0.01478844384352366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,2560,0.016528000434239704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,3584,0.033026665449142456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,4096,0.04602577620082431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,3072,0.029488000604841445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,3584,0.04105866617626614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,2048,0.014127110441525778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,2560,0.012621333201726278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,2560,0.025663110944959853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,1536,0.011539555258221097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,3072,0.03653600149684482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,2048,0.011424000064531961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,2560,0.031311111317740545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,1024,0.008988444175985124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,2048,0.021944888763957556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,1536,0.010626667075686984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,768,0.00739733295308219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,1536,0.01791200041770935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,2048,0.027271999253167048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,1024,0.009647111097971598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,512,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,1536,0.02312177750799391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,768,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,768,0.011959999799728394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,256,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,512,0.00868622213602066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,512,0.010049777726332346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,768,0.016591111818949383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,128,0.004190222256713443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,256,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,3072,0.01382844481203291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,512,0.014866666661368476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,64,0.0038364442686239877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,128,0.008384000096056197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,256,0.013170666992664337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,32,0.003966222206751506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,64,0.008295111358165741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,128,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,65536,0.26439377996656627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,1024,0.013806222213639153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,65536,0.17247733804914686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,1024,0.01887022289964888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,51200,0.16421600182851157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,65536,0.39651555485195583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,16384,0.06860444280836317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,51200,0.13724799950917563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,51200,0.3221537801954481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,128,0.00830577810605367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,16384,0.10299288564258152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,12288,0.04833155539300707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,16384,0.04688444402482775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,51200,0.43384443389044863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,32,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,16384,0.14736533164978027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,12288,0.03666755557060242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,12288,0.07710222403208415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,10240,0.0325911111301846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,12288,0.11282666524251302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,8192,0.03258133265707228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,10240,0.06516266531414457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,8192,0.02444533341460758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,10240,0.09576177597045898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,7168,0.028616888655556574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,8192,0.052821334865358144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,256,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,6144,0.023549333214759827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,8192,0.07878577709197998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,7168,0.021720889541837905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,65536,0.5495724678039551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,7168,0.046775112549463906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,5120,0.021239999267790053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,6144,0.0196782218085395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,7168,0.06943644417656793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,6144,0.04101599918471442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,4096,0.01814844376511044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,5120,0.03472800056139628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,10240,0.041794667641321816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,6144,0.061724444230397545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,3584,0.016145777371194627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,4096,0.015493333339691162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,5120,0.053189333942201406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,4096,0.029205332199732464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,3072,0.013958222336239286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,3584,0.014509333504570855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,3584,0.026492445005310908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,4096,0.044064889351526894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,2560,0.012208888928095499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,3072,0.013471110827393003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,3072,0.023345778385798138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,2048,0.010307555397351583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,3584,0.03930222325854831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,2560,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,1536,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,3072,0.03500888744990031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,2560,0.020632000433074105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,2048,0.011376889215575324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,1024,0.006139555739031897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,2560,0.030630220969518025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,1536,0.010438222024175854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,1536,0.01477866702609592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,2048,0.02642311155796051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,1024,0.00941422250535753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,1024,0.011776888536082374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,1536,0.022688888841205176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,5120,0.01737955543729994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,512,0.004511111312442356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,768,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,768,0.01015644437736935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,1024,0.018569777409235638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,256,0.003544888976547453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,512,0.008674666285514832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,512,0.008637333081828224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,768,0.016520889268981088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,128,0.003170666595300039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,256,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,256,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,64,0.0029253332565228143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,128,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,128,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,32,0.0031795555518733132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,128,0.010339555641015371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,64,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,2048,0.017482666505707633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,65536,0.19511466556125215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,32,0.007234666910436418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,768,0.0053102220926019884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,51200,0.1575440035925971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,65536,0.169250660472446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,16384,0.05152800016933017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,65536,0.34548266728719074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,51200,0.27409066094292533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,65536,0.4091155529022217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,16384,0.04615555538071526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,12288,0.03979555434650845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,51200,0.32452532980177135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,256,0.012872888810104795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,16384,0.11276533868577744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,10240,0.030498666895760432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,12288,0.035884443256590105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,12288,0.06849422057469685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,8192,0.024248000648286607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,10240,0.0317831105656094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,12288,0.08648000160853068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,10240,0.057392001152038574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,7168,0.022671999202834234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,8192,0.023697778582572937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,8192,0.0470142232047187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,10240,0.07293422354592217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,6144,0.0192275560564465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,51200,0.13896177874671087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,7168,0.021500445074505273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,8192,0.0610355536142985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,5120,0.01629688839117686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,6144,0.01958577831586202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,6144,0.036779556009504534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,16384,0.09038311243057251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,7168,0.05301422211858961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,512,0.014490667316648694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,4096,0.013651555611027611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,5120,0.017317333155208163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,5120,0.03158400124973721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,3584,0.012243555651770698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,4096,0.015243555108706156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,4096,0.026612444056404963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,5120,0.04109333289994134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,3072,0.010724444356229572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,3584,0.023584889041052923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,4096,0.034120000070995755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,3584,0.014383999837769402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,2560,0.00959822204377916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,3072,0.013415999710559845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,3072,0.021313778228229944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,3584,0.031063109636306763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,2048,0.008277333445019191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,2560,0.018744889232847426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,2560,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,2048,0.01585244470172458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,2048,0.011208000282446543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,2560,0.02438755498992072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,2048,0.02164622147878011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,6144,0.04760800136460198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,1024,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,1536,0.013365333278973898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,1536,0.010412444670995077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,1536,0.018526222970750596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,768,0.004857777721352047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,1024,0.009384888741705153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,1024,0.015602666470739575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,768,0.009383999639087254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,512,0.003960000144110786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,768,0.009015999734401703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,7168,0.04202666547563341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,768,0.014436443646748861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,256,0.0034968890249729156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,512,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,1536,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,128,0.0030906666070222855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,256,0.00829777783817715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,256,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,128,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,64,0.0028275555620590844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,1024,0.0107342220014996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,128,0.009131555755933126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,32,0.0028640000770489373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,64,0.00657066661450598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,65536,0.16489333576626247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,32,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,3072,0.02825511164135403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,51200,0.12968266010284424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,512,0.008629333641793992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,65536,0.12480000654856364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,65536,0.3022666772206624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,16384,0.0414373344845242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,512,0.012799999780125089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,51200,0.10009777545928955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,256,0.011729777687125735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,65536,0.4059884548187256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,51200,0.23690488603379992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,16384,0.07942755354775323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,12288,0.026888888743188646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,51200,0.3173271020253499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,10240,0.024596444434589807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,12288,0.059993777010175914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,10240,0.024578667349285547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,8192,0.019880889190567862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,10240,0.050733334488338895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,10240,0.0722115569644504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,8192,0.04160711169242859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,7168,0.01723022262255351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,8192,0.017808889349301655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,8192,0.05996088849173652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,128,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,16384,0.03533422284656101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,12288,0.031217777066760596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,7168,0.037047111325793795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,7168,0.016552888684802584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,5120,0.013419555293189155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,6144,0.014888889259762235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,7168,0.053247110711203686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,16384,0.11217599444919163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,6144,0.032631112469567194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,4096,0.011027555498811932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,5120,0.013575110998418597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,5120,0.02788089050187005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,12288,0.08558399809731378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,6144,0.047205332252714366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,4096,0.012044444680213928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,4096,0.02321333355373806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,5120,0.04001333316167196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,3072,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,3584,0.011399110986126794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,4096,0.03352266550064087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,2560,0.008656889200210571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,3584,0.03049422303835551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,3072,0.011013333168294696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,3072,0.01906488835811615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,2560,0.010286221901575724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,3072,0.027172444595230952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,2560,0.016881777180565726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,6144,0.015435554915004306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,2048,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,2560,0.024481778343518574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,1536,0.006492444624503453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,2048,0.014430221584108142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,2048,0.02143999934196472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,1536,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,3584,0.010281778044170803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,1024,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,1536,0.012120000190205045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,768,0.004569777597983678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,1024,0.008330666356616551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,1536,0.018404444058736164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,1024,0.009720000127951304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,3584,0.021104888783560857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,512,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,768,0.008240888516108194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,768,0.00871733327706655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,256,0.003181333343187968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,1024,0.015503999259736804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,512,0.007589333587222629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,512,0.007732444339328342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,768,0.013596444494194455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,2048,0.007671111159854465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,256,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,256,0.006686222222116258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,512,0.012303111453851065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,64,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,32,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,128,0.007325333025720384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,256,0.011414222419261932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,128,0.00608800020482805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,65536,0.10585688882403904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,64,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,51200,0.08560799890094334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,32,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,128,0.008625778059164682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,65536,0.12372799714406331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,16384,0.03077510992685954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,65536,0.2579057746463352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,51200,0.09983555475870769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,12288,0.02441066669093238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,51200,0.20318933327992758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,16384,0.034902221626705594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,16384,0.06766044431262545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,10240,0.02089599933889177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,65536,0.38987289534674746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,16384,0.10823021994696723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,12288,0.026558222042189703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,51200,0.3073555628458659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,12288,0.05168177684148153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,128,0.0027475555737813315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,10240,0.023525334066814844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,10240,0.04386755492952135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,7168,0.013633777697881063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,12288,0.08209066920810275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,8192,0.01758666667673323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,6144,0.012275555895434486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,8192,0.036144889063305326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,10240,0.06964000066121419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,7168,0.016155555844306946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,5120,0.010415111151006486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,7168,0.03232266505559286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,8192,0.05747377872467041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,6144,0.01500888831085629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,6144,0.0284222231970893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,4096,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,7168,0.051105777422587075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,5120,0.013491555220550962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,3584,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,6144,0.04459199971622891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,4096,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,4096,0.020573332905769348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,3072,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,5120,0.0377768874168396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,3584,0.011406222151385413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,3584,0.01863822175396813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,2560,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,3072,0.010816000401973724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,8192,0.016972444123691983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,3584,0.029282665914959375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,2048,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,2560,0.014880888991885714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,2560,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,3072,0.02648266653219859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,1536,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,2048,0.009732444253232744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,2048,0.012864888542228274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,2560,0.023144889209005568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,1024,0.0035004446076022256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,1536,0.010997333460383944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,1536,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,2048,0.020557334025700886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,768,0.003216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,1024,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,1024,0.00907733374171787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,1536,0.017887999614079792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,768,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,768,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,1024,0.014767999450365702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,512,0.0028275555620590844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,4096,0.03242399957444932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,256,0.0025351111673646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,512,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,3072,0.016869333055284288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,512,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,128,0.00221422231859631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,256,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,256,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,512,0.011767110890812345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,64,0.0021537778278191886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,128,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,32,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,256,0.010681777364677854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,65536,0.0664586689737108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,64,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,128,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,32,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,5120,0.02422844370206197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,51200,0.05347555544641283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,65536,0.12320178084903294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,65536,0.22147733635372588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,16384,0.01863555610179901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,51200,0.17300444179111055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,51200,0.09887199931674534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,768,0.01314133322901196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,12288,0.013444444371594323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,65536,0.38615555233425564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,16384,0.03401777810520596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,51200,0.3042622142367893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,10240,0.01183555523554484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,12288,0.02609955602222019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,12288,0.04455111092991299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,128,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,16384,0.10628622108035618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,8192,0.010472888747851053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,10240,0.023405333360036213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,12288,0.0813608898056878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,10240,0.037588444021013044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,7168,0.011704000333944956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,8192,0.01756888959142897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,8192,0.030900445249345567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,6144,0.010208889014191097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,10240,0.06875377893447876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,7168,0.016187555260128446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,8192,0.056905779573652476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,7168,0.02775466607676612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,5120,0.009005332986513773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,6144,0.014852444330851236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,4096,0.007701333198282454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,6144,0.024384000235133704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,5120,0.013267555170589022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,7168,0.050435556305779346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,3584,0.00703288863102595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,5120,0.021368000242445204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,6144,0.04394488864474826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,4096,0.012099555797047086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,16384,0.058168888092041016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,4096,0.018184888694021437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,3072,0.006320888797442119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,5120,0.03737066520584954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,3584,0.011533333195580376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,2560,0.005969777703285217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,3584,0.016557332542207506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,3072,0.010936000280910067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,2048,0.004805333498451444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,3072,0.015100444356600443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,2560,0.010341333018408881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,3584,0.029135111305448744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,2560,0.013450666434235044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,1536,0.004200889004601372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,3072,0.02609599961174859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,2048,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,2048,0.011644444531864591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,1024,0.0035013332962989807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,2560,0.022974222898483276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,1536,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,768,0.0032151111712058387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,1536,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,2048,0.019941333267423842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,1024,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,512,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,1024,0.008364444805516137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,1536,0.01754577789041731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,768,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,256,0.0025111111915773815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,768,0.0074746666683091064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,1024,0.014495111174053617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,512,0.007683555285135905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,128,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,512,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,768,0.013073777986897362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,256,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,256,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,64,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,32,0.002130666747689247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,128,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,4096,0.031553778383466936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,128,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,256,0.010374222364690568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,65536,0.06155021985371908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,51200,0.04950933323966133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,128,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,32,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,12288,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,65536,0.22027111053466797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,51200,0.17297066582573783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,10240,0.009975999593734741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,16384,0.05810666746563382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,12288,0.04450666573312548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,7168,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,10240,0.037237332926856145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,8192,0.03066844410366482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,6144,0.009762666291660732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,5120,0.008408889174461365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,7168,0.027501333090994094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,4096,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,512,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,3584,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,6144,0.02436888880199856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,5120,0.021029333273569744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,4096,0.01794399983353085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,16384,0.01519555515713162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,3584,0.016543111867374845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,2048,0.004530666602982415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,3072,0.014829332629839579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,1536,0.00388355553150177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,1024,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,2560,0.013460444079505073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,8192,0.008830221990744272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,2048,0.0114702218108707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,512,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,1536,0.010029333333174387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,256,0.0026266665922270883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,1024,0.008295999632941352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,128,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,768,0.007470221983061896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,64,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,512,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,32,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,256,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,128,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,65536,0.06306666798061795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,51200,0.048134220971001514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,64,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,3072,0.005741333381997214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,16384,0.013004444539546967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,32,0.005095110999213325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,12288,0.010065777434243096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,64,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,10240,0.009280000295903947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,2560,0.005042666776312722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,65536,0.21952799956003824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,16384,0.058005332946777344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,7168,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,12288,0.04455822375085619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,6144,0.0071928890215026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,10240,0.03752444518937005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,5120,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,8192,0.030472887886895075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,768,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,3584,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,7168,0.027313777142100867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,6144,0.02418311105834113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,3072,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,2560,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,4096,0.01796444422668881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,2048,0.004513777792453766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,3584,0.016533333394262526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,1536,0.0038737778862317405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,3072,0.014856888188256158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,1024,0.0031964443624019623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,2560,0.013460444079505073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,768,0.0028400001012616684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,2048,0.011424889167149862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,512,0.00254488881263468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,1536,0.010114666488435533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,256,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,1024,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,128,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,8192,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,768,0.0076044441925154785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,64,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,32,0.0018728888697094387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,512,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,51200,0.17270133230421278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,256,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,128,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,4096,0.008103999826643202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,64,0.0052266667286554975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,32,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,5120,0.02106844385464986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,16384,1.0715511110093858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,12288,2.670017878214518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,12288,0.8020995457967123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,12288,2.627478281656901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,16384,3.2238434685601125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,16384,3.5915857950846353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,10240,0.7040559980604383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,10240,2.0297173394097223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,51200,4.019989437527127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,10240,2.250457763671875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,8192,1.6564213434855144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,8192,0.5521137979295518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,8192,1.7095333735148113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,7168,0.4861484633551703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,6144,0.43066933419969344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,7168,1.4150622685750325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,6144,1.3468133078681097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,7168,1.937562730577257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,6144,1.2831306457519531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,5120,1.260706689622667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,5120,0.36925421820746523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,4096,0.3011191156175402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,4096,0.8927982118394641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,5120,1.053408940633138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,3584,0.811922656165229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,3584,0.2688773208194309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,4096,0.874293327331543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,3584,0.7296586566501193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,3072,0.6433128780788845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,3072,0.624060418870714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,3072,0.25037601259019643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,2560,0.5936542087131077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,2560,0.5202453401353624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,2048,0.17612977822621664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,2048,0.464030212826199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,1536,0.4200755490197076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,2048,0.4307288858625624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,1536,0.14047200149959987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,1024,0.2376240094502767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,1536,0.33369867006937665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,1024,0.11164177788628472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,768,0.19506222671932646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,51200,10.32577429877387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,512,0.1534808874130249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,768,0.09813600116305882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,1024,0.35563288794623477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,768,0.22813778453403047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,2560,0.21210932731628418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,256,0.08731022146013047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,128,0.07599111398061116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,512,0.17951022254096138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,256,0.09755555788675944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,64,0.06810488965776232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,128,0.08072711361779107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,256,0.1542853381898668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,32,0.0801431139310201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,512,0.09189511007732815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,128,0.14584710862901476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,64,0.1494026713901096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,32,0.150325338045756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,16384,2.487545861138238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,51200,3.2252142164442272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,51200,16.077701144748264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,65536,4.0806702507866754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,16384,0.9389822218153211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,16384,2.8996605343288846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,12288,0.6492942174275717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,12288,2.1478150685628257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,10240,1.6065617667304144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,12288,2.1811341179741754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,51200,7.943845960828994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,10240,0.5442969004313151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,8192,1.2731919818454318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,10240,1.8049528333875868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,8192,0.4288364516364203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,8192,1.311457739935981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,7168,0.3865342140197754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,7168,1.3763181898328993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,65536,10.111991034613714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,6144,0.3322693241967095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,7168,1.1420799891153972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,6144,0.9546719657050239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,5120,0.2841831048329671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,4096,0.6041342417399088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,4096,0.24585866928100586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,6144,1.0434541702270508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,4096,0.6338337792290581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,5120,0.8186364703708224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,3584,0.5650373564826118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,51200,11.665564643012154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,3584,0.21243733829922148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,3584,0.5662097930908203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,5120,0.9500124189588758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,3072,0.47897688547770184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,3072,0.1920480065875583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,2560,0.16664977868398032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,3072,0.4800764189826117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,2048,0.2978595627678765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,2560,0.43032267358567977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,2048,0.1375111076566908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,2560,0.4108702341715495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,1536,0.2514417701297336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,1024,0.1962568892372979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,1536,0.11141510804494222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,2048,0.38384800487094456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,768,0.13604977395799425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,1536,0.26229243808322483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,1024,0.08767555819617377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,512,0.10518844260109796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,1024,0.20082578394148085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,256,0.07946222358279757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,512,0.07194044192632039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,768,0.22078133953942192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,128,0.05694222450256348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,512,0.13604266113705105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,256,0.07311555412080553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,64,0.048482666412989296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,256,0.12193689081403945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,128,0.06167822413974338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,32,0.0506533318095737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,128,0.14357688691880968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,64,0.12341954973008896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,32,0.11943999926249187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,768,0.07891466882493761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,65536,1.320767084757487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,16384,0.8043431176079644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,51200,0.986088858710395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,16384,0.295580440097385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,65536,15.944542778862846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,12288,0.21119822396172416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,12288,0.8379991319444445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,51200,2.789087083604601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,16384,0.9594995710584852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,65536,3.2422498067220054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,12288,0.6898969014485677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,10240,0.19528444608052573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,10240,0.5479022132025825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,8192,0.45596175723605686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,8192,0.1473004420598348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,7168,0.3878222306569417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,7168,0.1594933271408081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,6144,0.3297475443945991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,51200,3.8223228454589844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,10240,0.7104631000094943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,7168,0.5219013426038954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,6144,0.12063466178046332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,5120,0.2630106608072917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,5120,0.09938666555616592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,4096,0.2225946585337321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,5120,0.28292621506585014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,4096,0.08469155761930679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,3584,0.19336533546447754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,4096,0.21790756119622123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,65536,4.907569037543403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,3072,0.16282578309377035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,3584,0.23057511117723253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,2560,0.1378035545349121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,3072,0.16581777731577554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,3072,0.06602222389645047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,2048,0.11000621981090969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,8192,0.4451671176486545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,1536,0.08414222134484185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,2048,0.049717333581712514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,2560,0.05773599942525228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,1024,0.0581715570555793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,1536,0.04733777708477444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,2048,0.11584799819522434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,768,0.04895200000868904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,1536,0.09104177686903213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,512,0.032552000549104475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,6144,0.3650355603959825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,768,0.03671111000908746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,768,0.056127998563978404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,256,0.026975999275843304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,512,0.029383109675513372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,128,0.021056888831986323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,512,0.04811733298831516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,3584,0.07266755236519708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,256,0.02683911058637831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,64,0.017687110437287223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,256,0.04326488905482822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,128,0.025756445195939805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,32,0.018254222141371835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,128,0.041336887412601046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,64,0.04218399855825636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,2560,0.14345599545372859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,32,0.041787554820378624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,1024,0.03438755538728502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,1024,0.07049866517384847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,51200,0.7526257832845052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,16384,0.6027448972066244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,65536,1.008839077419705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,16384,0.6884720060560438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,16384,0.21970310476091173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,12288,0.4889938036600749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,12288,0.1656915611690945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,51200,1.9475972917344835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,10240,0.42569422721862793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,12288,0.599305788675944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,10240,0.14071022139655218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,65536,2.450288984510634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,8192,0.32506577173868817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,8192,0.11410844326019287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,10240,0.4925209151373969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,7168,0.26511910226609975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,7168,0.10779199997584026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,8192,0.4085128837161594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,6144,0.26256712277730304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,7168,0.30186578962537974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,6144,0.089519997437795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,6144,0.2472462124294705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,5120,0.2086657815509372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,51200,3.0064746008978953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,4096,0.17248088783688018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,5120,0.07596800062391493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,4096,0.06849600209130181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,5120,0.22442756758795845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,3584,0.14867644839816624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,3072,0.1277351114484999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,4096,0.1624115573035346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,3584,0.05686489078733656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,3072,0.05626489056481255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,65536,3.668963538275825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,3584,0.15228711234198675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,2560,0.04542666673660278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,3072,0.12707822852664524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,2048,0.09256177478366429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,1536,0.06761244270536634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,2560,0.10713689194785224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,2048,0.03933866818745931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,1536,0.034078223837746516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,2048,0.09396088785595363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,768,0.03741155399216546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,1536,0.0718195570839776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,1024,0.026009778181711834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,512,0.02852355440457662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,1024,0.054083555936813354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,768,0.02464977734618717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,256,0.01902311046918233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,768,0.0444488889641232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,128,0.016353777713245816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,512,0.03576888971858554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,256,0.019711999429596793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,64,0.014072888427310519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,256,0.031960888041390315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,32,0.014116444521480136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,128,0.0195795562532213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,128,0.030856887499491375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,64,0.030293332205878362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,2560,0.10410488976372613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,32,0.03015466531117757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,1024,0.0451546675629086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,512,0.021929777330822412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,16384,0.5205901993645562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,65536,0.7931057612101237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,51200,0.6183884408738878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,16384,0.20740532875061035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,16384,0.553168879614936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,12288,0.4424168798658583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,65536,2.0145626068115234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,51200,1.7253138224283855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,12288,0.13781955507066515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,10240,0.34042933252122665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,12288,0.44921334584554035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,8192,0.2923591136932373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,10240,0.12514044178856745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,8192,0.09791288773218791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,7168,0.19509422779083252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,8192,0.2830471197764079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,7168,0.09085244602627224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,6144,0.21109778351253936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,51200,2.4684836069742837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,5120,0.18447466691335043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,6144,0.07475822501712374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,7168,0.3036195437113444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,6144,0.20853866471184623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,4096,0.12705066468980578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,5120,0.0717759993341234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,3584,0.1237857739130656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,5120,0.18316888809204102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,65536,3.07096438937717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,3072,0.09842489163080852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,4096,0.13987199465433756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,2560,0.08561066786448161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,3072,0.04682133264011807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,2048,0.06599822309282091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,3072,0.10994666814804077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,2560,0.04123199979464213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,2560,0.09222755829493205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,1536,0.05487733417087131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,2048,0.03347644540998671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,1024,0.03644444545110067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,10240,0.4991768731011285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,1536,0.02830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,768,0.027902222341961328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,1536,0.060589333375295006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,1024,0.022982221510675218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,512,0.019935111204783123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,1024,0.04629066586494446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,768,0.03854488995340135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,256,0.015571556157535978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,512,0.018795554836591084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,512,0.03040800160831875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,3584,0.05074577861362033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,256,0.017165333032608032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,3584,0.14107555813259548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,256,0.02683999968899621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,64,0.011582222249772815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,128,0.026007110873858135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,128,0.01719555589887831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,32,0.011567110816637674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,64,0.02608088817861345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,32,0.025455110602908667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,2048,0.07774844434526232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,768,0.02074666652414534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,4096,0.05537244346406725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,128,0.013528000149461957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,65536,0.6570115619235569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,16384,0.43737512164645725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,51200,0.5169617864820693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,51200,1.3408159679836695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,16384,0.15210399362776014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,65536,1.8625662061903212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,12288,0.3760862350463867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,16384,0.5065235561794704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,12288,0.1174515618218316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,10240,0.29534933302137584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,12288,0.34848088688320583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,10240,0.09984178013271755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,8192,0.2461635536617703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,8192,0.08059555292129517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,7168,0.18265510929955375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,10240,0.36472709973653156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,7168,0.08414577775531346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,51200,2.07279290093316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,65536,2.4141805436876083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,6144,0.16391911771562365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,5120,0.14880800247192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,8192,0.31307466824849445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,6144,0.06336177719963922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,4096,0.09791377517912124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,5120,0.05824089050292969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,3584,0.10437511073218451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,6144,0.1800631152258979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,4096,0.04558311237229241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,3072,0.08437422249052261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,5120,0.17474399672614205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,3584,0.04163733455869887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,2560,0.0669084456231859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,4096,0.1253653367360433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,3072,0.03667199942800734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,3584,0.10411377747853597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,1536,0.04419377777311537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,2560,0.03297422329584757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,3072,0.09018044339285956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,2560,0.08335200283262466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,2048,0.028736889362335205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,2048,0.0641991098721822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,1536,0.023295111126369897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,768,0.02408711115519206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,1024,0.020026667250527274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,1536,0.05048533280690511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,512,0.016893333858913846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,1024,0.038543111748165555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,768,0.01834755473666721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,256,0.013789333403110504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,768,0.03218044506178962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,512,0.01622577839427524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,128,0.010679110884666443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,512,0.025391111771265667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,7168,0.19901600148942736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,64,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,32,0.009900444083743626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,128,0.015320888823933072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,128,0.02169688873820835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,64,0.02162666618824005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,32,0.021635555558734473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,2048,0.059398220645056836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,1024,0.028379556205537584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,65536,0.6117271317376031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,256,0.01514755520555708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,256,0.0225600004196167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,51200,0.4769626723395453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,51200,1.1479724248250325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,16384,0.13892355230119494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,16384,0.46129157808091903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,12288,0.31354933314853245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,16384,0.4007120132446289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,12288,0.1063199970457289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,10240,0.26742400063408744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,10240,0.08883288833830093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,12288,0.3158773316277398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,8192,0.2308755450778537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,10240,0.3138924439748128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,51200,1.7924630906846788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,65536,2.1695662604437933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,7168,0.1625137726465861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,8192,0.07181421915690105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,6144,0.14812000592549643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,8192,0.2792106734381782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,5120,0.1412408881717258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,6144,0.056663109196556934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,7168,0.1920915577146742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,4096,0.093176888095008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,5120,0.04905688762664795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,6144,0.15976622369554308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,3584,0.07547110981411405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,65536,1.4985307057698567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,4096,0.04121866822242737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,5120,0.14147555828094482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,3072,0.07679555813471477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,4096,0.10133866469065349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,3584,0.03773333297835456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,2560,0.06263822317123413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,3072,0.03589244352446662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,3584,0.09102577633327907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,2048,0.04884888728459676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,3072,0.07898844612969293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,2560,0.030594666798909504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,1536,0.038916445440716214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,2560,0.06764888763427734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,1024,0.02677244444688161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,2048,0.025246222813924152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,768,0.02064622276359134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,1536,0.021943110558721755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,1024,0.017906667457686532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,1536,0.04492444462246365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,512,0.016072000066439312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,768,0.01680355601840549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,1024,0.03485777642991807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,256,0.011677333050303988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,7168,0.06380622254477607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,512,0.01513777838812934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,768,0.028597331709331934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,128,0.010291555689440833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,512,0.022974222898483276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,64,0.008656000097592672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,32,0.008993777963850234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,256,0.020289777053727042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,128,0.01960444450378418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,64,0.01961688862906562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,32,0.019612444771660697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,2048,0.05798044469621447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,65536,0.5255049069722494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,65536,1.1964631610446506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,256,0.01427288850148519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,51200,0.8842888938056098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,128,0.014155555102560254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,16384,0.31775199042426217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,12288,0.236191987991333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,51200,0.41747376653883195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,16384,0.3517431153191461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,10240,0.1967866685655382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,12288,0.10063999891281128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,12288,0.25815023316277397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,65536,1.795861350165473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,8192,0.1590577761332194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,10240,0.07863466607199775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,10240,0.29312266243828666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,8192,0.06389422549141778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,7168,0.05654578076468574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,6144,0.1290000014834934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,51200,1.5791688495212133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,7168,0.1872417794333564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,5120,0.10487022002538045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,4096,0.087391111585829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,6144,0.13228710492451987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,5120,0.04353866643375821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,5120,0.14209866523742676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,3584,0.07553955581453112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,4096,0.03632266653908624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,16384,0.12559910615285239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,3072,0.05603733327653673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,4096,0.09491199917263454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,3584,0.03335111008750068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,2560,0.05227999885876974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,3584,0.0884702205657959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,3072,0.030113776524861652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,3072,0.06985866361194186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,2048,0.04244266615973579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,2560,0.02684533264901903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,1536,0.03361422154638503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,2560,0.05968622366587321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,2048,0.023294222023752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,7168,0.1426533328162299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,1536,0.01982222166326311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,2048,0.05070755547947354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,768,0.02074399921629164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,8192,0.17816710472106934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,1024,0.016992888516849942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,512,0.01514755520555708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,6144,0.04910133282343546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,768,0.015400000744395785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,768,0.025610667135980394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,512,0.014197332991494073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,128,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,512,0.020590222544140287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,256,0.013122667041089801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,256,0.018248000078731112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,128,0.013530666629473368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,32,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,128,0.01755911111831665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,64,0.01756000022093455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,32,0.017525333497259352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,1024,0.024521778027216595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,1536,0.04028088847796122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,1024,0.030530667967266504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,256,0.012712000144852532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,65536,0.45036178165011936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,64,0.007751999629868402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,65536,1.0706710815429688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,16384,0.34259645144144696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,16384,0.10565244489245945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,51200,0.847110218471951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,16384,0.29503822326660156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,12288,0.24873688485887316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,65536,1.5513057708740234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,12288,0.08245422442754109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,10240,0.17263644271426729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,12288,0.2215475506252713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,8192,0.15078311496310765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,10240,0.07231822278764513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,51200,1.1678942574395073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,8192,0.055604444609748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,10240,0.22011378076341417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,8192,0.147709329922994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,7168,0.13756444719102648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,6144,0.04292710953288608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,6144,0.11768889427185059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,5120,0.08335110876295303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,51200,0.36221154530843097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,4096,0.06572622060775757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,5120,0.03760000069936117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,3584,0.05971822473737928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,4096,0.03137866655985514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,5120,0.09566844171947902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,3072,0.050372444921069674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,4096,0.07682844665315416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,2560,0.04457155532307095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,3072,0.02565866708755493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,3072,0.06098666456010607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,2048,0.033203555477990046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,2560,0.023224888576401606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,7168,0.13002755906846789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,2560,0.05215822325812446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,1536,0.02791644467247857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,7168,0.05041777756479052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,2048,0.020257777637905545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,1024,0.01945955554644267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,2048,0.04411733481619093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,1536,0.017510222064124215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,768,0.015426667200194465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,1024,0.014871999621391296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,512,0.011388444238238864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,768,0.01371911085314221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,768,0.022750222020679053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,256,0.00868711123863856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,512,0.017990221579869587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,3584,0.02773777809407976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,128,0.007679999702506595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,256,0.01183999992079205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,256,0.016083555089102853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,64,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,128,0.011873777541849347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,32,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,128,0.015485333071814643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,64,0.015511999527613321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,32,0.01531377765867445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,1536,0.03503733211093479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,1024,0.026537777649031744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,512,0.012692444854312472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,65536,0.3941484557257758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,65536,0.8504888746473523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,3584,0.06885866986380683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,51200,0.6821484565734863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,51200,0.3159795602162679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,16384,0.21874133745829263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,16384,0.09280622005462646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,12288,0.1678417788611518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,6144,0.10732622279061212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,12288,0.07451822360356648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,65536,1.1927724414401584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,10240,0.16134488582611084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,51200,0.9422808753119575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,8192,0.13922754923502603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,8192,0.049421333604388766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,7168,0.12040621704525417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,10240,0.21837777561611602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,8192,0.1307555569542779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,6144,0.09824799829059178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,7168,0.04361688759591845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,7168,0.11138933234744602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,5120,0.08142222298516168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,6144,0.036794665786955096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,4096,0.06556889083650377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,6144,0.10191022025214301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,5120,0.0335440006521013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,5120,0.07965066697862414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,3584,0.06047022342681885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,4096,0.028484443823496502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,3072,0.050365332100126475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,4096,0.06541777981652154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,3584,0.024491555160946313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,2560,0.038703110482957624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,3584,0.058041777875688344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,3072,0.022106667359670002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,2048,0.02990844514634874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,2560,0.019893333315849304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,16384,0.3515671094258626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,3072,0.051065779394573636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,1536,0.02458222210407257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,2560,0.04381600022315979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,10240,0.06087999873691135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,2048,0.01761155492729611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,1024,0.017587555779351127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,2048,0.03696088989575704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,1536,0.015251555376582675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,1536,0.029637333419587877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,1024,0.013311111264758639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,512,0.01069777790043089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,1024,0.022388445006476507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,768,0.012424888710180918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,256,0.007589333587222629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,768,0.019268444842762418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,128,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,512,0.015589333242840238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,256,0.010933332973056369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,256,0.01386044422785441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,128,0.010848888920413123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,32,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,64,0.013395555317401886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,32,0.013406222065289816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,12288,0.2005546622806125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,768,0.014183999763594734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,512,0.011717333561844297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,65536,0.7829528914557563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,65536,0.34521155887179905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,51200,0.5342426829867893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,16384,0.21090755197736952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,128,0.013441777891582914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,65536,1.010943094889323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,12288,0.153674668735928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,16384,0.2492213249206543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,12288,0.06868710782792833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,12288,0.16434755590226915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,51200,0.8005493481953939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,8192,0.10342311196857029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,10240,0.05708533525466919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,7168,0.09622844722535874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,8192,0.043549332353803843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,8192,0.11175555653042263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,7168,0.04018133216434055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,64,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,7168,0.09941422277026707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,5120,0.06701155503590901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,6144,0.03416533271471659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,6144,0.08552089002397324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,51200,0.3125671015845405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,4096,0.048826667997572154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,5120,0.030200001266267564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,3584,0.04650577902793884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,10240,0.12935466236538357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,4096,0.02606577840116289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,3072,0.035748445325427584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,3584,0.023781332704755995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,3584,0.0521395570702023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,3072,0.021431111627154883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,2560,0.029222223493787978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,10240,0.14101155598958334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,3072,0.046625776423348315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,2560,0.019662222928471036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,6144,0.07465155257119073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,1536,0.02072000006834666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,2048,0.01680888897842831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,2048,0.03396711084577773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,1024,0.014702222413486905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,1536,0.014470222923490735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,16384,0.08716711070802476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,1536,0.02730133301681942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,768,0.011807111402352652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,1024,0.01275644451379776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,512,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,768,0.01220266686545478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,768,0.017942221628295053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,4096,0.05806755357318454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,256,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,512,0.011386666860845355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,512,0.014460444450378418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,128,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,256,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,2048,0.026695999834272597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,64,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,256,0.012879110872745514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,32,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,128,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,128,0.012416000167528788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,2560,0.04048088855213589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,64,0.01239466667175293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,32,0.012340444657537671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,65536,0.2968631055619982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,65536,0.5804799927605523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,1024,0.020868443780475195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,51200,0.23453156153361002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,5120,0.07150489091873169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,51200,0.5181040234035915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,16384,0.15645244386461046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,12288,0.12563378281063503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,16384,0.07147377729415894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,10240,0.09941866662767197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,12288,0.06271644433339436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,16384,0.21112888389163545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,12288,0.14320000012715658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,65536,0.889144049750434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,8192,0.08163999848895602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,51200,0.7037635379367404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,10240,0.053611556688944496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,7168,0.07120888763003878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,10240,0.1223848925696479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,6144,0.062315556738111705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,8192,0.03771022293302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,7168,0.03554844525125291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,5120,0.053054223457972206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,8192,0.10141333606508042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,6144,0.03071555495262146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,7168,0.09063466389973958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,5120,0.027387556102540758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,3584,0.037724445263544716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,6144,0.07732977469762166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,5120,0.06388622522354126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,3072,0.031063109636306763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,4096,0.053539554278055825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,3584,0.02096533278624217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,2560,0.02645422187116411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,3072,0.01923022170861562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,3584,0.04751466711362203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,2048,0.022688888841205176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,3072,0.042111999458736844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,2560,0.01756444407833947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,1536,0.018256889449225534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,2048,0.015488000379668342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,2560,0.036722666687435575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,1024,0.013797333670987023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,1536,0.013929777675204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,2048,0.03087288803524441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,768,0.01185955521133211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,1024,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,1536,0.02473955518669552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,512,0.008850666383902231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,768,0.011258666714032492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,256,0.0063475556671619415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,768,0.016190222567982145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,512,0.010572444233629437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,128,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,512,0.013460444079505073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,4096,0.041927112473381885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,256,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,64,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,32,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,256,0.01183644433816274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,128,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,4096,0.022952889402707417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,128,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,64,0.011386666860845355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,32,0.01142133358452055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,65536,0.26831645435757107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,65536,0.5511253145005968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,51200,0.3946213457319472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,51200,0.21410489082336426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,16384,0.14627556006113687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,12288,0.10974222421646118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,16384,0.07662400272157457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,1024,0.018924444913864136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,10240,0.08538222312927246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,12288,0.053173333406448364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,65536,0.7357955508761935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,51200,0.5783066749572754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,8192,0.07519644498825073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,12288,0.12777511278788248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,7168,0.06644355588489108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,8192,0.0364951127105289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,6144,0.05714133050706652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,10240,0.11568978097703721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,7168,0.03272977802488539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,5120,0.04633422361479866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,8192,0.08744355705049302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,7168,0.07824711004892985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,6144,0.029193778832753498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,5120,0.02604444490538703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,3584,0.03262133399645487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,4096,0.021887999441888597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,5120,0.05655733082029554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,3072,0.026206221845414903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,3584,0.020277332928445604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,3584,0.04179911149872673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,2560,0.023856888214747112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,3072,0.01835911141501533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,3072,0.03722488880157471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,2048,0.01928711103068458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,2560,0.01699555582470364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,16384,0.17115467124515107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,1536,0.014814222852389017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,2560,0.032235556178622775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,2048,0.015192889504962497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,1024,0.010777778095669217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,2048,0.026974222726292078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,1536,0.01256622208489312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,1536,0.021890666749742296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,768,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,4096,0.03530399998029073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,1024,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,1024,0.01698933376206292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,6144,0.06802666849560209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,768,0.010454222559928894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,256,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,768,0.014477334088749357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,512,0.009758222434255812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,4096,0.0468355549706353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,128,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,512,0.011905777785513135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,64,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,256,0.009593778186374241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,32,0.004620444443490771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,128,0.009710222482681274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,256,0.01072799993885888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,128,0.01032977799574534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,64,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,32,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,10240,0.04608355628119575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,65536,0.37551376554701066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,65536,0.2342382272084554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,51200,0.2948719925350613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,512,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,65536,0.5841715600755479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,51200,0.18832266330718994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,16384,0.06272444460127088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,12288,0.08890666564305623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,16384,0.14429244730207655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,10240,0.07031910949283175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,51200,0.4603653483920627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,12288,0.045132444964514844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,8192,0.05753866831461588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,12288,0.11022400193744236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,10240,0.04004888733228048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,8192,0.030866665972603693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,10240,0.08998222483528985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,6144,0.04225066635343763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,8192,0.07407733466890123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,5120,0.03510844376352098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,7168,0.06592266427146064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,6144,0.024680889315075342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,4096,0.026641777820057336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,6144,0.05630399783452352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,5120,0.022042666872342426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,3584,0.024717332588301763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,5120,0.04743466774622599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,4096,0.018767111831241183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,3072,0.020642666353119742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,16384,0.1074097752571106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,3584,0.017335999343130324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,2560,0.017850667238235474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,3584,0.03560622202025519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,3072,0.01590044465329912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,2048,0.0166320006052653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,2560,0.014834667245546976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,3072,0.031331555710898504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,1536,0.013268444273206921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,2560,0.02730577852990892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,2048,0.01367644468943278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,1024,0.009707555174827576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,1536,0.012124444047609964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,2048,0.02299022177855174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,768,0.007839110990365347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,7168,0.04768888817893135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,1536,0.018982221682866413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,512,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,1024,0.014797333214018079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,768,0.010050666828950247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,256,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,7168,0.02828977836502923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,512,0.00958577791849772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,768,0.01275644451379776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,128,0.004181333300140169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,512,0.01054044481780794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,256,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,32,0.0038897775941424896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,128,0.00904088881280687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,256,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,128,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,64,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,32,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,4096,0.03934399949179755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,65536,0.34300888909233945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,65536,0.17506755722893608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,16384,0.08975644244088067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,51200,0.2535395622253418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,51200,0.14013866583506265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,12288,0.06633155875735812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,65536,0.47708442476060653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,10240,0.05003199974695841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,12288,0.04168711105982462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,51200,0.3747733434041341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,16384,0.12095022201538086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,8192,0.04300533400641548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,12288,0.08996088637246026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,10240,0.03750666644838121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,7168,0.03858933183881972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,8192,0.025016888976097107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,64,0.0038364442686239877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,8192,0.06203022268083361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,6144,0.03215644425816006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,7168,0.02273422148492601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,5120,0.027311111489931744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,7168,0.05453333258628845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,6144,0.02055377761522929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,4096,0.02201777696609497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,6144,0.04776800009939405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,5120,0.01798577772246467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,3584,0.019762666689025033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,4096,0.015782222151756287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,5120,0.04074044360054864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,3072,0.017044444878896076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,4096,0.034146666526794434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,3584,0.014829332629839579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,16384,0.04639288783073425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,2560,0.014828445182906257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,3584,0.030794666873084173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,3072,0.013611555927329592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,2048,0.01253600004646513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,3072,0.02681511143843333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,1536,0.009911999934249455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,2560,0.01260355528857973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,2048,0.011347555451922946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,2560,0.023398222194777593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,1024,0.007727111379305522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,1536,0.01039644413524204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,2048,0.019945777124828763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,768,0.006755555669466655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,1536,0.0165928883685006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,1024,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,512,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,768,0.009316444396972656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,1024,0.012862222062216865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,256,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,768,0.011384000380833944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,512,0.008681777450773451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,128,0.0038373333712418876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,512,0.009656888743241629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,64,0.00349866681628757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,128,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,256,0.008376888930797577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,32,0.003567999849716822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,128,0.00833511104186376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,64,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,65536,0.2127093341615465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,32,0.00816000004609426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,1024,0.010677333507272931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,51200,0.15738133589426676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,16384,0.060589333375295006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,65536,0.16477778222825792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,51200,0.13089244895511204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,12288,0.044180442889531456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,10240,0.07586933506859674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,16384,0.04474133253097534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,10240,0.03432800041304694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,16384,0.10014044576221043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,12288,0.03520888752407498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,8192,0.02753066685464647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,10240,0.03092977735731337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,12288,0.07301421960194905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,7168,0.025847999585999384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,10240,0.06135911411709256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,256,0.00832266691658232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,8192,0.024432889289326135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,6144,0.02183288832505544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,5120,0.018075555562973022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,7168,0.021477333373493616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,6144,0.01925511161486308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,7168,0.0447964436478085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,4096,0.015095111396577625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,5120,0.017275555266274344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,6144,0.039507554637061224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,3584,0.013785777820481194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,5120,0.03348977698220147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,4096,0.01516177753607432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,3072,0.011758222348160215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,4096,0.02790044413672553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,3584,0.014180445008807711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,2560,0.010100444157918295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,3072,0.013122667041089801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,3584,0.025116443634033203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,2048,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,65536,0.37401154306199813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,3072,0.022503111097547743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,2560,0.012067555553383298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,1536,0.007170666423108842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,2560,0.019835554891162448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,1024,0.005653333332803514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,2048,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,768,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,1536,0.010099555883142684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,1024,0.009343110852771336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,1536,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,512,0.004224888980388641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,768,0.008984888593355814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,1024,0.011337777806652917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,256,0.00350577798154619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,512,0.008336000144481659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,768,0.010009778042634329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,8192,0.050199998749627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,512,0.008309333688682979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,256,0.008016000191370646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,64,0.002847111059559716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,256,0.007502222226725684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,128,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,32,0.002926222152180142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,128,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,64,0.00700444479783376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,32,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,51200,0.14746222231123182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,65536,0.16211288505130345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,51200,0.29190754890441895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,65536,0.3258062203725179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,2048,0.017050666941536795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,16384,0.052813334597481616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,12288,0.03887733485963609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,51200,0.12863643964131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,16384,0.044059554735819496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,10240,0.028894222444958154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,51200,0.2589600086212158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,16384,0.08572089009814793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,128,0.003156444471743372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,8192,0.02475111186504364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,12288,0.03412177827623155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,12288,0.06461422310935126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,7168,0.0214720004134708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,10240,0.03062844607565138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,6144,0.01808355583084954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,10240,0.05477244324154324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,7168,0.020963556236690942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,8192,0.04457777738571167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,5120,0.016188444362746347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,7168,0.0398942232131958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,4096,0.013645333548386892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,6144,0.019152889649073284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,5120,0.016985777351591323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,3584,0.010967111421955956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,4096,0.0151235560576121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,5120,0.029856890439987183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,3072,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,4096,0.025069332785076563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,3584,0.014106666048367819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,2560,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,3584,0.022669333550665114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,3072,0.013046222428480784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,3072,0.020473778247833252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,2048,0.007141333487298753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,2560,0.012127111355463663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,1536,0.006052444378534953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,2560,0.01792266633775499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,2048,0.011019555230935415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,1024,0.004868444469239977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,2048,0.01550044450494978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,1536,0.010112000008424124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,768,0.004224888980388641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,1024,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,1536,0.012824888858530255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,8192,0.022890667120615642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,512,0.0037084445357322693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,1024,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,768,0.009000889129108852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,512,0.008359111017651027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,768,0.009091555244392818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,65536,0.2016755607393053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,128,0.0031537777847713898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,256,0.008204444415039485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,64,0.0028826666788922418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,6144,0.035339554150899254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,32,0.0028906667398081887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,256,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,128,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,128,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,65536,0.159844438234965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,64,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,32,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,51200,0.1277795500225491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,16384,0.03683822353680929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,51200,0.09565155373679267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,65536,0.28460176785786945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,16384,0.033350222640567355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,12288,0.02646488944689433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,10240,0.020223110914230347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,51200,0.22348088688320586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,16384,0.07440266344282362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,12288,0.026012443833880957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,8192,0.016142222616407607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,10240,0.023192889160580103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,12288,0.05670844184027778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,7168,0.014195554786258273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,8192,0.01736000014675988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,10240,0.04811999864048428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,6144,0.01237066669596566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,8192,0.03949244485961066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,7168,0.01612622208065457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,5120,0.010827555424637265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,512,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,6144,0.014706666270891825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,7168,0.035347554418775774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,4096,0.0092275556590822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,5120,0.013167111410035027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,3584,0.00830488900343577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,4096,0.011726222104496427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,5120,0.02649066680007511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,3072,0.007583111524581909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,4096,0.022296888960732356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,65536,0.11867288748423259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,3584,0.01112711098459032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,2560,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,3584,0.02032088902261522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,3072,0.010692444112565784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,2048,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,3072,0.018238221605618794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,1536,0.004871110949251386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,2560,0.016136889656384785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,2048,0.013936888840463428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,1536,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,1024,0.004084444294373195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,1536,0.011709333293967776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,768,0.003556444413132138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,1024,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,1024,0.009360888765917884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,512,0.003195555466744635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,768,0.0081395556529363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,768,0.008621333373917473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,6144,0.03084889054298401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,256,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,512,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,128,0.0024844445288181305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,512,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,256,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,64,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,128,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,32,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,2560,0.010105777945783403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,2048,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,65536,0.08490400181876288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,64,0.005976888868543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,32,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,51200,0.06528800063663058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,256,0.0031831111345026228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,65536,0.11815199587080215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,16384,0.023823999696307715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,12288,0.017981333865059745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,51200,0.09514133135477702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,16384,0.03231111168861389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,65536,0.24250398741828072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,51200,0.19104533725314668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,10240,0.018408000469207764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,12288,0.024901333782407973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,16384,0.0636684430970086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,12288,0.04854399959246317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,10240,0.022504000200165644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,7168,0.013797333670987023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,256,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,8192,0.017088888419999015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,10240,0.041454222467210554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,6144,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,5120,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,8192,0.03413510984844632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,7168,0.015935111376974318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,6144,0.014517333772447375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,4096,0.010176888770527309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,7168,0.03039111031426324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,3584,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,5120,0.013227555486891003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,6144,0.026704889204767015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,3072,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,5120,0.02316177719169193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,4096,0.011782222323947482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,2560,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,3584,0.011315555208259158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,4096,0.019539554913838703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,2048,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,3584,0.017907554904619854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,2560,0.010097777677906884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,1536,0.0042239998777707415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,2048,0.009671111073758867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,2048,0.012429333395428128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,1024,0.003519111209445529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,1536,0.009039999710188972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,8192,0.015444444285498725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,1536,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,768,0.003216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,1024,0.008361777497662438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,512,0.0028728888266616394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,768,0.008072888685597314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,1024,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,256,0.002534222271707323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,768,0.007805333369308048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,512,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,128,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,128,0.0023235556566052968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,512,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,256,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,64,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,32,0.0022115556316243275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,128,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,128,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,3072,0.010493333141009012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,64,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,65536,0.06447555621465047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,3072,0.0159262220064799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,51200,0.05202755663130018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,32,0.005495999836259418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,2560,0.014167999227841696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,16384,0.014531556102964612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,65536,0.11767555607689752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,12288,0.012064889073371887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,51200,0.09486755397584702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,16384,0.032242665688196816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,65536,0.20486933655209014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,10240,0.011237333218256632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,51200,0.16068355242411295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,16384,0.05430577860938179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,12288,0.024365334047211543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,8192,0.00998755544424057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,7168,0.01125600023402108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,10240,0.02269422180122799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,8192,0.01720266706413693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,10240,0.03532266616821289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,6144,0.011569778124491373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,8192,0.02906222144762675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,7168,0.015764445066452026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,5120,0.00869599978129069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,256,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,7168,0.025881777207056682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,6144,0.014489778214030795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,4096,0.007632888853549957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,5120,0.013711999687883588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,6144,0.02295733326011234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,3584,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,3072,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,4096,0.011897777517636618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,3584,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,2560,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,3072,0.010696888797812991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,3584,0.01573511130279965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,3072,0.01420088940196567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,2560,0.010055111514197456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,2560,0.012720000412729053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,1536,0.0041999999019834726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,2048,0.009663999908500248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,2048,0.01128266668981976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,1024,0.0033742222521040174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,12288,0.0418177776866489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,1536,0.008974221845467886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,768,0.0031866667171319327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,1024,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,512,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,1024,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,768,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,256,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,768,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,512,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,128,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,512,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,256,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,5120,0.019967110620604623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,64,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,32,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,128,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,4096,0.017088888419999015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,128,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,2048,0.004588444613748127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,64,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,51200,0.04548799991607666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,16384,0.012683555483818054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,32,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,12288,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,10240,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,65536,0.20390133062998453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,8192,0.009298666483826106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,16384,0.05428533421622383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,51200,0.16051822238498265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,12288,0.04111288984616598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,7168,0.00979644474056032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,6144,0.008411555654472774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,1536,0.009676444033781687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,5120,0.008267555799749162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,4096,0.006903111106819577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,8192,0.028633776638242934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,7168,0.025773333178626165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,3584,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,6144,0.02293333411216736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,3072,0.0057591112951437635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,2560,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,5120,0.0199297782447603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,2048,0.00453244439429707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,4096,0.01718133356836107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,1536,0.0038951109680864546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,256,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,1024,0.003176888864901331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,3072,0.014214222629865011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,768,0.002875555513633622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,2560,0.012702222499582501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,65536,0.05925510989295112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,512,0.002543111021320025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,2048,0.011301333705584208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,256,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,1536,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,128,0.0021271111650599376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,1024,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,64,0.002011555557449659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,768,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,32,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,256,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,128,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,65536,0.0565226674079895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,64,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,51200,0.04453955425156487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,10240,0.03469511204295688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,32,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,16384,0.012486221889654795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,12288,0.010197333163685268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,10240,0.009191110730171204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,65536,0.2040168841679891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,12288,0.041265779071384005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,8192,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,7168,0.007727111379305522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,3584,0.015642666154437594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,6144,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,8192,0.02857511242230733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,5120,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,6144,0.023003555006451074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,4096,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,5120,0.02000977761215634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,3584,0.007754666937722101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,4096,0.017049777838918898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,3072,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,3584,0.01573511130279965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,2560,0.006151999864313338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,16384,0.054153776831097074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,51200,0.16017156177096897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,2048,0.00453599997692638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,3072,0.014144889182514615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,1536,0.003871110992299186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,10240,0.034852445125579834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,1024,0.003170666595300039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,768,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,2560,0.012712889247470431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,512,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,2048,0.011341333389282227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,1536,0.009671111073758867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,256,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,1024,0.008031110796663497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,7168,0.025773333178626165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,128,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,64,0.0018088887963030075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,768,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,32,0.0018639999131361644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,512,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,512,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,256,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,128,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,64,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,32,0.005044444567627377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,16384,1.0224328570895724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,16384,3.214903089735243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,16384,3.6588889227973094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,12288,2.662130567762587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,16384,3.0742354922824435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,51200,3.8145857916937937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,12288,0.7381697760687934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,12288,2.507063971625434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,10240,0.6266809039645725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,10240,2.298484378390842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,12288,2.263967090182834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,10240,2.3074728647867837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,8192,0.5129280090332031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,8192,1.8483272128634984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,8192,1.59552976820204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,7168,1.6676408979627821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,8192,1.573635525173611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,10240,1.9550524817572699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,51200,10.601869371202257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,7168,1.3934035831027562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,7168,0.46207910113864475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,7168,1.3699493408203125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,6144,0.4066782262590196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,6144,1.1635280185275607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,6144,1.2054568396674263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,5120,0.3400257693396674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,5120,1.2552870644463432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,4096,0.7796168857150607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,5120,1.0110196007622612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,6144,1.078181372748481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,4096,0.30340621206495494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,5120,0.9611333211263021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,4096,0.8184977637396919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,3584,0.6575324270460341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,3584,0.2559448877970378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,3584,0.7179013358222114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,4096,0.7379697693718804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,3072,0.5135199758741591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,3072,0.2231431007385254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,3584,0.6533377965291342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,3072,0.7139377593994141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,2560,0.5131404664781358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,2560,0.5595911343892416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,3072,0.5767573250664605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,2560,0.19864888985951742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,2048,0.4652293523152669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,2048,0.16773422559102377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,2560,0.47364266713460285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,1536,0.350524451997545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,2048,0.41737156444125706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,2048,0.3965555561913385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,1536,0.32729511790805393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,1536,0.1482586728201972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,51200,10.391955905490452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,768,0.1786657836702135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,1536,0.3068133460150825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,1024,0.2647555669148763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,1024,0.22761776712205675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,51200,15.042700873480904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,768,0.09419111410776775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,768,0.26678400569491917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,512,0.12010310755835639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,256,0.09031822284062703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,768,0.18944888644748262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,512,0.09797422091166179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,512,0.16973066329956055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,128,0.07105511426925659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,1024,0.20421955320570204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,256,0.07599822017881605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,512,0.15449955728318956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,256,0.15820088651445177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,64,0.06489600075615777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,32,0.06558222240871854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,256,0.12063022454579671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,128,0.07169600327809651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,1024,0.11364621586269802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,128,0.091666665342119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,128,0.14355733659532335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,32,0.1470844480726454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,64,0.14535644319322374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,65536,3.863447825113932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,51200,2.7937901814778647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,51200,7.987006293402778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,16384,2.929632822672526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,16384,3.0959947374131946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,16384,0.7903297742207845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,65536,10.243185255262587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,12288,1.9917227427164714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,12288,0.5907920201619467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,51200,11.80101097954644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,16384,2.485207027859158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,51200,8.202732510036892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,12288,2.1782845391167536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,12288,1.693076451619466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,10240,0.5446062088012695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,10240,1.6150070826212566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,10240,1.6081004672580297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,8192,0.40441422992282444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,8192,1.4364008373684352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,10240,1.5361493428548176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,8192,1.3118764029608834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,7168,1.103680928548177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,65536,15.354800754123263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,7168,0.3616666793823242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,65536,10.366009182400173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,7168,1.138519075181749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,8192,1.1502186457316081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,6144,0.9312541749742297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,6144,0.31620534261067706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,6144,0.9334203932020398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,5120,0.26775911119249135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,7168,1.0355315738254123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,5120,0.8616711298624674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,5120,0.7762373288472494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,4096,0.6645208994547526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,6144,0.8446862432691785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,4096,0.22408888075086805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,4096,0.6110906600952148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,5120,0.7192177772521973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,3584,0.5848248799641927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,3584,0.19762844509548613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,3584,0.5383129119873047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,4096,0.5705786810980903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,3072,0.5549128850301107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,3072,0.17502933078342012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,3072,0.46640708711412215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,2560,0.423509332868788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,3584,0.5065884590148926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,2560,0.18789154953426787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,3072,0.4361697832743327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,2048,0.35615377955966526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,2560,0.3986079957750108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,2048,0.13158844576941595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,1536,0.281040006213718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,1024,0.1573057836956448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,1536,0.10795822408464219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,1536,0.26803821987575954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,2048,0.3165511025322808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,768,0.117740445666843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,1536,0.2532702287038167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,1024,0.2245404455396864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,1024,0.08567911386489868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,1024,0.18227999740176729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,512,0.1209377712673611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,768,0.07634933127297296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,768,0.1605786614947849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,256,0.08425955639945136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,768,0.15117511484358045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,512,0.13385511769188776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,512,0.12316266695658366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,256,0.12025422520107693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,256,0.060158219602372914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,128,0.055456890000237354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,2560,0.3734453254275852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,64,0.04602400130695767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,256,0.0966977808210585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,128,0.05788266658782959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,32,0.059474666913350425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,128,0.11409778065151638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,128,0.07378666930728488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,64,0.11490844355689155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,32,0.11560266547732884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,512,0.06780711147520277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,2048,0.3260355525546604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,65536,1.1970044242011177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,51200,0.8834026654561361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,51200,2.587898678249783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,16384,1.0899546941121419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,65536,3.287706587049696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,16384,0.26068001323276097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,16384,1.0722897847493489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,12288,0.6398604181077745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,16384,0.7650648752848307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,12288,0.19750932852427164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,12288,0.6388017866346571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,51200,4.076325310601129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,51200,2.6818603939480252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,65536,4.722925398084852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,12288,0.5539804564581977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,10240,0.5179608662923177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,10240,0.18136000633239746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,8192,0.4275217851003011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,8192,0.13932977782355413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,10240,0.5453724331325954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,65536,3.457029342651367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,7168,0.37746487723456484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,8192,0.44562578201293945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,10240,0.461075570848253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,7168,0.12290222114986843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,8192,0.3750240008036296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,6144,0.31924888822767467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,6144,0.10979466968112522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,5120,0.2430880069732666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,5120,0.09493422508239746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,7168,0.3333457840813531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,4096,0.19369599554273817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,5120,0.26111822658114964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,6144,0.2835386594136556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,3584,0.16471821731991237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,4096,0.08049511247211032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,4096,0.20968800120883516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,5120,0.3339431020948622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,3584,0.18278666337331137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,3072,0.1538906627231174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,4096,0.19946132765875924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,2560,0.1314648919635349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,3072,0.06370844443639119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,3584,0.1742995580037435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,3072,0.22749688890245226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,7168,0.4148346583048503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,2048,0.09829066859351264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,2560,0.05528888768619961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,3072,0.152800891134474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,6144,0.3742435508304172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,1536,0.07604889074961345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,2048,0.04713244570626152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,2560,0.1328995492723253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,2048,0.11243289046817356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,1024,0.05734399954477946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,1536,0.039600888888041176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,1536,0.08990666601392959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,2048,0.11071377992630005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,768,0.03874311182234023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,1024,0.033004442850748696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,1536,0.08984800179799397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,1024,0.06898311111662123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,512,0.03381866547796462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,768,0.030464887619018555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,768,0.05527911252445645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,1024,0.07011022170384724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,256,0.022085333863894146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,512,0.04757155643569099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,512,0.027845333019892376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,128,0.01991555591424306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,256,0.042633778519100614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,256,0.025518221987618342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,512,0.049485332436031766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,2560,0.14034222231970891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,64,0.01681688924630483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,256,0.038712888956069946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,32,0.017549332645204332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,128,0.040673779116736516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,128,0.024600888291994732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,64,0.040941334433025785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,128,0.03015022145377265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,32,0.04121689001719157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,3584,0.06934044758478801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,768,0.0596604413456387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,65536,0.8602399826049805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,51200,2.006323496500651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,16384,0.6825075679355197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,65536,2.5924835205078125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,51200,0.776519086625841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,16384,0.2026613288455539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,16384,0.6765279769897461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,12288,0.6036737759908041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,51200,2.7790525224473743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,16384,0.5631457964579264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,12288,0.18914311461978486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,10240,0.37480444378323025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,12288,0.47687286800808376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,65536,3.818322499593099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,51200,2.003650665283203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,10240,0.13085156016879612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,8192,0.3208017879062229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,12288,0.4170817799038357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,65536,2.5947653452555337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,10240,0.41649688614739316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,8192,0.10524000061882867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,8192,0.34380356470743817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,7168,0.26519645584954155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,7168,0.09576177597045898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,10240,0.35850932863023544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,6144,0.23071021503872344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,8192,0.327601777182685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,7168,0.29964176813761395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,5120,0.19360621770222983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,6144,0.2415093315972222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,7168,0.27362489700317383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,4096,0.15485778119828966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,5120,0.07183911403020223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,5120,0.214319109916687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,6144,0.22267021073235405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,4096,0.0708284444279141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,3584,0.12164266904195149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,4096,0.16143200132581922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,5120,0.18396711349487305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,3072,0.11416978306240505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,3584,0.057687997817993164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,2560,0.09997333420647515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,3584,0.14251111613379583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,3072,0.05015644431114197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,3584,0.13521244790818956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,3072,0.1228053304884169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,2560,0.04389422138532003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,2048,0.08408355712890625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,3072,0.11993955241309272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,2560,0.10841155714458889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,1536,0.061816891034444175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,6144,0.08437866634792751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,2048,0.03758488761054145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,2560,0.10444800059000652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,1536,0.030582222673628066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,1536,0.06962400012546115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,2048,0.08958933088514541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,768,0.03235288792186313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,1024,0.05314844515588549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,1536,0.07205866442786323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,1024,0.024859555893474158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,512,0.022817777262793645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,768,0.02293866707219018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,1024,0.055607113573286265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,768,0.042690667841169566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,256,0.02481511069668664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,4096,0.15296444627973768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,512,0.020787555310461257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,512,0.03505955470932855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,768,0.04802577694257101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,128,0.020928889513015747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,256,0.018626666731304593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,256,0.030920889642503526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,64,0.013491555220550962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,128,0.02993511160214742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,256,0.03183555603027344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,2048,0.08733066585328843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,128,0.02439733346303304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,64,0.029847999413808186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,32,0.029571556382709082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,512,0.03991733325852288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,128,0.018742221924993727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,32,0.013481777575280933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,1024,0.04537599947717455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,65536,0.7689938015407987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,16384,0.6319813198513454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,65536,2.1217901441786027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,51200,0.5941217740376791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,16384,0.19871200455559623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,51200,1.8987901475694444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,16384,0.5867359903123643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,12288,0.4115777810414632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,16384,0.4794524510701497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,65536,3.1518063015407987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,12288,0.13255822658538818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,65536,2.213006125556098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,10240,0.3531351089477539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,12288,0.5509271091885036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,51200,2.5999599032931857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,12288,0.3640248775482178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,51200,1.7317084206475153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,10240,0.12273600366380479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,10240,0.3422275649176703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,8192,0.2730426788330078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,8192,0.10415733522839016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,10240,0.3115359942118327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,8192,0.3318471113840739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,6144,0.17624799410502115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,7168,0.09352533022562663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,7168,0.2347102165222168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,8192,0.2463680108388265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,5120,0.16530311107635498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,7168,0.22941954930623373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,6144,0.07382133271959093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,6144,0.20612000094519722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,4096,0.1382604440053304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,5120,0.062182221147749156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,5120,0.16810311211480033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,6144,0.18722489145067003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,3584,0.1234097745683458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,4096,0.05213155680232578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,4096,0.1360702249738905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,5120,0.16370756096310085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,3072,0.09609422418806289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,3584,0.046446220742331616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,4096,0.1318933301501804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,7168,0.24676799774169922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,3584,0.11969155735439724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,2560,0.08014133241441515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,3072,0.04114222195413377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,3072,0.10684888892703587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,2048,0.06552888949712117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,3584,0.11815822124481201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,2560,0.03707022137112088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,1536,0.0541928874121772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,2560,0.09021866983837551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,3072,0.10308177603615655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,2048,0.03200088938077291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,1024,0.039806223577923246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,2048,0.07457511292563544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,1536,0.0254737784465154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,2560,0.09035199880599976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,1536,0.05958222018347847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,1024,0.021456888980335657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,2048,0.07668622334798177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,1024,0.045985778172810875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,512,0.022506665852334764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,1536,0.06292800108591716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,768,0.019761777586407132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,768,0.03685955537690057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,256,0.018192888961897958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,512,0.01813066667980618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,512,0.030127998855378892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,768,0.042262223031785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,256,0.016560888952679105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,512,0.03565066721704271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,256,0.026372444298532274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,64,0.01424800025092231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,256,0.02847733431392246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,128,0.025802666942278545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,32,0.014502222339312235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,64,0.025429333249727886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,128,0.02160888910293579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,32,0.02546399997340308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,768,0.030602667066786025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,1024,0.04895288745562235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,128,0.015276443627145557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,65536,0.6117608812120225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,128,0.016709332664807636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,65536,1.5823876063028972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,51200,1.2890702353583443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,16384,0.4112844467163086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,51200,0.4910586675008138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,16384,0.14341066943274602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,16384,0.4679688877529568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,12288,0.33955910470750594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,16384,0.40591732660929364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,65536,2.349654303656684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,12288,0.126111110051473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,51200,2.051243464152018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,10240,0.27409688631693524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,65536,1.8584363725450304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,51200,1.3688240051269531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,12288,0.3004373444451226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,10240,0.28617599275377065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,8192,0.23512532975938585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,8192,0.07705066601435344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,10240,0.25849421819051105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,7168,0.166883561346266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,8192,0.25996266471015084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,6144,0.1459475490781996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,7168,0.06792355246014066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,7168,0.2176835536956787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,8192,0.21908533573150635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,5120,0.13758844799465603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,6144,0.06055911382039388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,12288,0.3393075466156006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,6144,0.17058222823672822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,7168,0.18051555421617296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,4096,0.11819377210405137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,5120,0.05165333218044705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,6144,0.15916444195641413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,5120,0.14623467127482095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,3584,0.08460444211959839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,10240,0.10433955325020684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,4096,0.04367555512322319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,4096,0.11109689209196304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,5120,0.13538044028811985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,3072,0.08130311303668551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,3584,0.03786311215824551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,4096,0.11240711477067734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,2560,0.06645155615276761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,3072,0.034303999609417386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,3072,0.08654044734107123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,2560,0.031066666046778362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,2048,0.04965066578653124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,3584,0.10055555237664117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,3072,0.08835466702779134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,1536,0.04203111264440748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,2048,0.02643644478585985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,2560,0.07641955878999498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,2048,0.062195552719963916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,1024,0.03014311194419861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,1536,0.02176888949341244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,2048,0.06520977947447035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,1536,0.05407911207940844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,1024,0.018326222896575928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,1024,0.03798755672242906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,768,0.022097777989175584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,1024,0.04247733288341098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,768,0.016912889149453905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,3584,0.09909511274761623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,512,0.017542221479945712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,768,0.031390223238203264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,256,0.01275199982855055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,512,0.015964445140626695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,768,0.03659644391801622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,512,0.025125333004527624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,2560,0.07454666826460096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,128,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,256,0.014581332604090372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,512,0.030167112747828167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,256,0.02200266718864441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,64,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,32,0.010832889212502373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,128,0.014384888940387301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,256,0.02397955622937944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,128,0.021311110920376245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,1536,0.049089776145087353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,64,0.021306667062971327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,32,0.02127733329931895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,128,0.018593778212865193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,51200,0.41725865999857586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,65536,0.5353768666585287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,65536,1.3560507032606337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,16384,0.36904178725348574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,51200,1.3222213321261935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,16384,0.46309688356187606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,16384,0.14613866806030273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,12288,0.32206932703653973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,51200,1.7232915030585394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,65536,2.060323503282335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,51200,1.257314682006836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,12288,0.114738663037618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,16384,0.36855377091301816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,12288,0.29208800527784556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,10240,0.22956977950202095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,8192,0.19195822874704996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,12288,0.2764577865600586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,10240,0.09021688832177056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,65536,1.6435510847303603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,7168,0.16494311226738825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,8192,0.06805066929923163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,6144,0.14251289102766249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,7168,0.060398221015930176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,8192,0.20361333423190645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,7168,0.1725262271033393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,5120,0.11485066678788926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,7168,0.18308799796634248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,6144,0.05376355515586006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,6144,0.1468248897128635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,5120,0.04597244328922696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,4096,0.0933297806315952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,6144,0.14565066496531168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,5120,0.12352800369262695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,3584,0.07542133331298828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,4096,0.038182222180896334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,5120,0.12543733914693198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,4096,0.10104355547163223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,3072,0.0673875543806288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,3584,0.0350195566813151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,10240,0.2533795568678114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,8192,0.19344888793097603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,3584,0.08825688891940647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,4096,0.10116444693671332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,2560,0.05000977714856466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,10240,0.23493422402275932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,3072,0.030986666679382324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,3584,0.09275733100043403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,2048,0.04857688811090258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,2560,0.06703111198213366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,3072,0.08045155472225614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,1536,0.03644711110326979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,2048,0.02421955598725213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,2560,0.0701448917388916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,1536,0.02054666644997067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,1024,0.025384000606007043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,1536,0.04425422350565592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,1536,0.04940711127387153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,1024,0.03425688876046075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,768,0.019276445110638935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,1024,0.017596445149845548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,1024,0.038776887787712946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,3072,0.07807288567225139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,768,0.016414221790101793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,512,0.014927110738224454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,768,0.028026666906144884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,2560,0.028083556228213843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,256,0.010833777487277985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,512,0.014468444718254937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,2048,0.05559200048446655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,512,0.02259822189807892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,128,0.01015022231472863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,256,0.019948444432682462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,512,0.027672888504134283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,128,0.019222221440739103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,256,0.021984888447655573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,64,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,32,0.008767999708652496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,128,0.01754488878779941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,64,0.019309333629078336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,32,0.019139554765489366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,768,0.03328888946109348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,2048,0.0603804455863105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,128,0.013781333135233985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,65536,1.3386995527479384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,65536,0.49736711714002824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,16384,0.3727848794725206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,51200,0.9186888800726997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,51200,0.3890417681799994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,256,0.013436444103717804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,16384,0.12057333522372776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,65536,1.4135679668850367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,16384,0.37419999970330137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,12288,0.2562391228146023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,51200,1.3170453177558052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,12288,0.08863377571105957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,16384,0.3285155561235216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,10240,0.2136231131023831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,51200,1.0626746283637154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,12288,0.2591466638776991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,8192,0.15657956070370144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,10240,0.07691111167271932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,10240,0.21537600623236763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,12288,0.25143554475572377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,8192,0.06324978007210626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,8192,0.18323111534118652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,65536,1.8202614254421656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,10240,0.2056302229563395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,7168,0.05470044414202372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,8192,0.16929422484503853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,7168,0.1503946648703681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,5120,0.1072124441464742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,7168,0.148561782307095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,6144,0.12803644604153103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,4096,0.08173866404427423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,5120,0.04098755452367995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,5120,0.10869599713219537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,6144,0.12961333327823216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,3584,0.07615911298327975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,4096,0.0345164438088735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,5120,0.11143822140163845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,4096,0.08704177538553874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,3072,0.06180444690916273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,3584,0.03124444352255927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,4096,0.09080266952514648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,3584,0.0774115522702535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,7168,0.1654133399327596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,2560,0.05059911145104302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,3072,0.027727999620967444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,6144,0.13507733080122206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,3584,0.08301955461502075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,3072,0.06789066394170125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,2048,0.04591022266281975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,2560,0.02535199953450097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,2560,0.058607108063168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,1536,0.034232000509897866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,6144,0.04771911104520162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,2048,0.021664889322386846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,1024,0.024933333198229473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,2560,0.06353777647018433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,2048,0.048920889695485435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,1536,0.01828888886504703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,1536,0.039170665873421565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,2048,0.05441866649521721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,768,0.01976444489426083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,1024,0.015506666567590503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,1024,0.030447999636332195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,1536,0.04464711083306206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,512,0.017467555072572496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,768,0.01459022197458479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,768,0.025060445070266724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,256,0.013714666995737286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,512,0.013437333206335703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,512,0.020389333367347717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,768,0.029684444268544514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,128,0.011516444385051727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,512,0.025335111551814612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,256,0.017901332841979135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,256,0.012619555824332766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,64,0.010663111176755695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,128,0.012703999876976013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,128,0.017215111189418368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,32,0.010809777511490716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,256,0.020962667134073045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,64,0.017234666479958426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,128,0.01688799924320645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,32,0.017216889394654166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,1024,0.03580088747872247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,65536,0.4424773322211371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,51200,0.762390242682563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,65536,1.2101768917507596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,3072,0.07422311438454522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,16384,0.2840862274169922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,51200,0.3450933297475179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,65536,1.5224435594346788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,12288,0.20848445097605386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,16384,0.11011022329330444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,16384,0.42921333842807347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,51200,1.1197342342800565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,65536,1.1806959576076932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,12288,0.08358399735556708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,16384,0.2840720017751058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,51200,0.9416444566514758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,12288,0.2333537737528483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,8192,0.13911733362409803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,10240,0.07240621911154853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,12288,0.23012177149454752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,8192,0.05363288852903578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,7168,0.1303688950008816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,8192,0.15827377637227377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,10240,0.2041893270280626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,7168,0.04753955536418491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,6144,0.10674844185511272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,8192,0.15022755993737116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,7168,0.1297475496927897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,5120,0.0795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,6144,0.04236266679233975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,6144,0.11287555429670547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,4096,0.06782133049435086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,5120,0.03633244501219855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,5120,0.09383377763960098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,6144,0.11527200539906819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,3584,0.061350219779544406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,4096,0.030576000610987347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,5120,0.09878222147623698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,10240,0.21168977684444854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,4096,0.076200889216529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,3072,0.048400001393424145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,3584,0.02796799937884013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,10240,0.18537510765923393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,3584,0.06709244516160753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,4096,0.08176088995403714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,2560,0.04299288988113403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,3072,0.025435555312368605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,3072,0.059007108211517334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,3584,0.07481155792872111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,2560,0.022666666242811415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,3072,0.06508888800938924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,2560,0.051065779394573636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,1536,0.02548800077703264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,2048,0.04275200102064344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,2048,0.019832889238993328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,2560,0.05674044291178385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,1024,0.01908533275127411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,1536,0.017172444197866652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,1536,0.034522665871514216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,2048,0.050028443336486816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,768,0.015343111422326831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,1024,0.026753778258959453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,1024,0.014711111783981323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,1536,0.040503998597462974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,512,0.011398221883508893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,768,0.022338666849666174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,768,0.013584888643688627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,1024,0.03215377860599094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,256,0.008610666626029545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,512,0.017863111363516915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,512,0.012695999609099494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,768,0.026946667167875502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,128,0.007349333001507654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,256,0.011776000261306763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,512,0.022679110368092854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,2048,0.032722665203942194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,64,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,128,0.015240889456537036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,128,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,32,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,128,0.015338665909237333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,7168,0.13240799638960096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,32,0.015173332558737861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,65536,0.3729066583845351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,65536,0.9164426591661242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,256,0.015857777661747403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,256,0.019312888383865356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,65536,1.2026888529459636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,64,0.015167999598715039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,65536,1.020467546251085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,16384,0.26228533850775826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,51200,0.2925191190507677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,51200,0.9079484939575195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,12288,0.19261956214904785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,16384,0.2405190997653537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,16384,0.09172088570064968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,51200,0.8920017878214518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,12288,0.06507022513283624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,10240,0.1650595532523261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,51200,0.7902773221333822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,16384,0.255112886428833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,10240,0.15399911668565539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,10240,0.057641777727339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,8192,0.1356755495071411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,12288,0.1969306601418389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,7168,0.11567910512288411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,8192,0.04547199938032362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,8192,0.12309955226050483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,6144,0.09381955862045288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,10240,0.15967022048102483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,7168,0.12845333417256674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,8192,0.13064711623721653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,5120,0.07716444465849134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,6144,0.03548444310824076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,7168,0.11599644025166829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,6144,0.09327644771999782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,4096,0.05996266338560316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,5120,0.07762489053938124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,6144,0.10087377495235866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,12288,0.24560266070895723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,3584,0.05379022161165873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,4096,0.026192888617515564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,5120,0.08667111396789551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,3072,0.04349244303173489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,3584,0.023938667443063524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,4096,0.0722551080915663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,3584,0.05689422289530436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,2560,0.0361697773138682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,3072,0.02199288871553209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,3072,0.05050933361053467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,3584,0.06492533286412557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,7168,0.04024799995952182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,2048,0.0306000014146169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,2560,0.01976088848378923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,2560,0.0436115562915802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,3072,0.05741600195566813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,1536,0.02306755549377865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,2048,0.017269333203633625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,2048,0.0369120008415646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,1024,0.016322667400042217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,2560,0.05040977729691399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,2048,0.04370133413208855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,1536,0.02937777837117513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,768,0.013088000317414602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,4096,0.06463822391298082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,1024,0.013056889176368713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,1536,0.03597333365016513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,1024,0.02251822253068288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,512,0.010592000352011787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,768,0.01240355521440506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,768,0.019024888674418133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,1024,0.028202666176689997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,256,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,512,0.015476443701320224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,512,0.011729777687125735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,128,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,768,0.02420444455411699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,256,0.013793778088357715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,256,0.010910222099887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,512,0.020582222276263766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,64,0.007368000017272101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,32,0.0074417781498697065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,128,0.010713777608341642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,128,0.013209777573744455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,64,0.01310755560795466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,128,0.014348444011476306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,5120,0.03111999895837572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,1536,0.01484266585773892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,65536,0.5929777887132432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,65536,0.30502578947279185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,51200,0.4572693506876628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,16384,0.17127911249796549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,51200,0.2389813264211019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,65536,1.0420381757948134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,256,0.01718488832314809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,51200,0.7969324323866102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,16384,0.23152621587117514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,65536,0.8560329013400607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,12288,0.13719377252790663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,32,0.01313688854376475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,16384,0.07393066750632392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,10240,0.12684622075822619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,12288,0.05918400155173408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,51200,0.672386646270752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,12288,0.16340711381700304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,8192,0.09289511044820149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,10240,0.1337617768181695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,10240,0.05016266637378269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,12288,0.17580444282955596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,7168,0.0867040024863349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,8192,0.04296177625656128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,10240,0.13886666297912598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,8192,0.10930399762259589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,7168,0.03676355547375149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,6144,0.07404355870352851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,7168,0.09780444701512654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,8192,0.11314488781823052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,5120,0.05797511339187622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,6144,0.031298667192459106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,7168,0.10115022129482693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,6144,0.08355822165807088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,4096,0.04662044511901008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,5120,0.027383999692069158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,5120,0.07025600141949125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,6144,0.08834399779637654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,3584,0.0392577780617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,4096,0.057992001374562584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,16384,0.23274400499131945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,3072,0.0365573333369361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,3584,0.021645334031846788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,4096,0.06435822116004096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,3584,0.05217511124081082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,2560,0.02737599942419264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,3072,0.019719110594855416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,3072,0.046463999483320445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,2560,0.018183999591403537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,2560,0.03996444410747952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,2048,0.024362666739357844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,2560,0.04536177714665731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,2048,0.03404266635576884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,1536,0.01867377758026123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,2048,0.016029333074887592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,2048,0.03881333271662394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,4096,0.023430221610599097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,5120,0.07514578104019165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,1024,0.013986666997273764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,1536,0.027162666122118633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,1536,0.014226666755146451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,768,0.011725333001878528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,1024,0.01237155579858356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,1024,0.020623111062579684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,1536,0.03267644511328803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,3584,0.05693778064515856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,512,0.008808000220192803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,768,0.01757777730623881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,768,0.011378666592968835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,1024,0.025422222084469263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,256,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,3072,0.052004443274603955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,512,0.014359111587206522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,128,0.006092444476154115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,768,0.02162933349609375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,256,0.010067555639478896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,64,0.005547555370463266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,256,0.012741333080662621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,512,0.018703111343913607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,32,0.00573511090543535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,128,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,128,0.012323555847009024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,256,0.016542222764756944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,64,0.012135110795497894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,32,0.012115555504957834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,128,0.013743999931547375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,65536,0.28960710101657444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,51200,0.45322667227851016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,16384,0.16329422261979845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,51200,0.2285146713256836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,512,0.010737777584128909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,65536,0.8703262011210123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,51200,0.6771590974595812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,12288,0.12567467159695095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,16384,0.07016355461544461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,51200,0.6474142074584961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,16384,0.19262666172451445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,65536,0.8947173224555122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,12288,0.05525333351559109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,12288,0.14531555440690783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,16384,0.2168311145570543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,8192,0.08603377474678887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,10240,0.04795733425352308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,10240,0.12114132775200738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,12288,0.16636444462670222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,65536,0.5787004364861382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,7168,0.07002577516767713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,8192,0.03733511103524102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,10240,0.1430471075905694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,6144,0.06023377842373318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,7168,0.03437422381507026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,8192,0.11203111542595758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,7168,0.09947644339667426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,6144,0.030060443613264296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,5120,0.054937776592042714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,6144,0.07483377721574572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,6144,0.08784621953964233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,5120,0.02648977769745721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,4096,0.043974220752716064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,5120,0.06291911337110731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,4096,0.022288888692855835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,3584,0.037598222494125366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,4096,0.06303289201524523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,8192,0.09757866462071736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,3584,0.0457893345091078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,3584,0.020640000700950623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,3584,0.05702400207519531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,3072,0.040722668170928955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,2560,0.02772533396879832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,3072,0.018936889039145578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,10240,0.11143022113376194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,2560,0.035424000687069364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,2560,0.01721600029203627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,3072,0.051113777690463595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,4096,0.05117600162823995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,2560,0.04459999998410543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,2048,0.015468445089128284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,1536,0.019069333871205647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,2048,0.029554665088653564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,7168,0.08595110972722371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,2048,0.03825777769088745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,1536,0.0241333329015308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,1536,0.031698667340808444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,1024,0.018547554810841877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,768,0.01129244433508979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,1024,0.012087110843923358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,1024,0.024752888414594863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,768,0.01614755557643043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,768,0.011365333365069495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,512,0.008489777644475302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,5120,0.07516800032721625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,768,0.021670222282409668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,256,0.00674222202764617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,512,0.010490666660997601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,512,0.013012444807423485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,3072,0.03250044584274292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,256,0.009986666341622671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,512,0.01828799976242913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,1024,0.014311111635631986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,256,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,64,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,128,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,256,0.015519110692871941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,32,0.005884444548024072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,128,0.011150221857759686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,64,0.01108266661564509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,128,0.012445333103338877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,32,0.01108088923825158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,65536,0.46221690707736546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,2048,0.02408444384733836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,65536,0.2649119959937202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,51200,0.3819439940982395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,16384,0.14221066898769802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,1536,0.013906665974193148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,65536,0.7494551340738932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,51200,0.21079199843936494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,12288,0.10264622502856785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,16384,0.0651351081000434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,65536,0.7001084221733941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,51200,0.5953840149773492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,16384,0.17163733641306558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,10240,0.09369155433442856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,51200,0.5393777953253852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,12288,0.054302222198910184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,16384,0.17851022879282633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,8192,0.07840889030032687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,10240,0.04713777701059977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,10240,0.10503111282984416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,12288,0.13753778404659694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,7168,0.06641333632998996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,8192,0.03775999943415324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,8192,0.08476355340745713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,10240,0.11567644278208415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,6144,0.0514977773030599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,7168,0.03086755673090617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,8192,0.09586133559544881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,7168,0.07512355513042875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,5120,0.045075555642445884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,128,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,6144,0.028209778997633193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,6144,0.06604533063040839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,4096,0.03690133492151896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,7168,0.0849555532137553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,5120,0.02420622275935279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,5120,0.05538577834765116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,3584,0.029853334029515583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,5120,0.06452266375223796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,4096,0.04535733328925239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,12288,0.127691560321384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,3584,0.04082666503058539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,4096,0.05321333474583096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,3584,0.018256889449225534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,3584,0.04801244537035624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,2560,0.022251556317011516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,3072,0.016750221451123554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,3072,0.036178665028678045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,2560,0.01569422251648373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,3072,0.04277333286073473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,2560,0.03148622314135233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,2560,0.03830844495031569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,2048,0.026330666409598455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,6144,0.07566399706734551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,1536,0.014689778288205465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,2048,0.014307555225160388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,4096,0.0203493336836497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,2048,0.0322106679280599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,1024,0.011022222538789114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,1536,0.012418666647540199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,3072,0.02829066581196255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,768,0.009177777502271865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,1024,0.010701333483060202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,1024,0.016535111599498324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,1536,0.026783999469545152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,512,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,1024,0.02166844407717387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,768,0.010393777655230628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,768,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,256,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,512,0.011706666813956367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,512,0.009670221971140968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,768,0.019190222024917603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,128,0.005031110925806893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,256,0.009424000150627559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,256,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,512,0.016492444607946608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,64,0.00452888881166776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,128,0.010021333065297868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,128,0.009082666701740688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,32,0.00462311092350218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,256,0.014840889308187695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,64,0.010069333016872406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,1536,0.021361778179804485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,32,0.010240889257854885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,65536,0.4368577798207601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,65536,0.21806577841440836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,65536,0.5824604564242893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,2048,0.018985778093338013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,16384,0.12966222233242458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,51200,0.172233780225118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,51200,0.4537537892659505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,16384,0.05710488557815552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,16384,0.14243377579583064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,65536,0.678433789147271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,128,0.012061333490742577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,51200,0.5292391247219509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,12288,0.04453600115246243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,16384,0.1754080057144165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,10240,0.07354399893018934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,12288,0.1074977782037523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,8192,0.05854044357935587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,10240,0.03827644387880961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,12288,0.13608800040351018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,51200,0.34538933965894913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,7168,0.05135644475618998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,8192,0.030191110240088567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,8192,0.07353955507278442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,6144,0.04245866669548882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,10240,0.11414044433169895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,7168,0.027415111660957336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,8192,0.09345510933134292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,5120,0.03504088852140639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,7168,0.06487644380993314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,6144,0.02432800001568264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,6144,0.056278222137027316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,7168,0.08308000034756131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,4096,0.02753244505988227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,12288,0.09489155477947658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,5120,0.021564443906148274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,3584,0.025604445073339675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,6144,0.07313333617316352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,4096,0.018624888526068795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,4096,0.039228445953793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,5120,0.06362844175762601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,3072,0.02199111051029629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,3584,0.01711466742886437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,3584,0.03542044427659776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,2560,0.019888000355826486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,4096,0.053039110369152494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,3072,0.015786666009161208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,10240,0.08984711435106064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,3584,0.0478168891535865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,3072,0.03137777911292182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,2048,0.016359110673268635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,2560,0.01459377838505639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,2560,0.027071999178992376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,1536,0.013007111019558378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,2048,0.013461333182122974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,3072,0.042352000872294106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,2048,0.023257777094841003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,2560,0.03678933448261685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,1024,0.009375111096435124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,1536,0.012150222228633033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,1536,0.0189057770702574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,2048,0.03135822216669718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,768,0.00776800016562144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,1024,0.010676444404655032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,1536,0.02618933386272854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,1024,0.01482222146458096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,512,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,5120,0.04701155424118042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,768,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,256,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,768,0.01275199982855055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,1024,0.02134311033619775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,512,0.009639110830095079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,128,0.004834666848182678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,512,0.01055466632048289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,768,0.018565333551830716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,256,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,64,0.004541333350870344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,256,0.009364444348547194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,128,0.008969777988062965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,512,0.015882665912310284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,32,0.004531555705600315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,128,0.009304000271691216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,256,0.014146667387750415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,64,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,128,0.011904888682895236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,32,0.009107555780145857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,65536,0.2783999972873264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,65536,0.17373511526319715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,51200,0.22142133447859022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,51200,0.1372399992412991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,65536,0.4755484263102214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,12288,0.062365333239237465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,51200,0.37285778257581925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,16384,0.046775112549463906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,16384,0.11877777841356064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,65536,0.5324719746907552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,10240,0.05613599883185493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,12288,0.0362737774848938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,51200,0.41635645760430234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,12288,0.09011733531951904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,16384,0.14317244953579372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,8192,0.04165244433614943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,10240,0.03222222129503886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,10240,0.07610755496554904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,8192,0.024481778343518574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,12288,0.11059555742475723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,8192,0.06157066424687704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,10240,0.09293244282404582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,6144,0.028821332587136164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,7168,0.022335999541812476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,8192,0.07663822174072266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,5120,0.027274666561020747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,6144,0.020145777199003432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,16384,0.08383911185794407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,6144,0.04710844490263197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,7168,0.06709066364500257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,4096,0.021599110629823472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,5120,0.017901332841979135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,5120,0.04049688908788893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,3584,0.019066666563351948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,4096,0.033936887979507446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,4096,0.015626667274369132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,5120,0.05234666665395101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,3072,0.016943999462657504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,4096,0.044072889619403414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,3584,0.030685332086351182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,7168,0.03700088792377048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,2560,0.014700444208251106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,3072,0.01344622257683012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,3584,0.03832711113823785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,2560,0.012421333127551608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,7168,0.05473866727617052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,2048,0.012523555921183692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,2560,0.023349333140585158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,2048,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,1536,0.009919111099508073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,2048,0.01995644470055898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,6144,0.060216890441046826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,2048,0.02630133264594608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,1024,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,1536,0.016545777519543965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,1536,0.010407999985747868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,3584,0.014536889062987434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,1024,0.009648000200589498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,1536,0.022693332698610093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,768,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,1024,0.012888000243239932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,512,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,3072,0.027173333697848853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,768,0.009162666896979014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,1024,0.018823999497625563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,768,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,256,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,512,0.009497777455382878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,3072,0.034633778863483004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,768,0.016498666670587327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,128,0.00387999994887246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,256,0.008423111504978603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,256,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,2560,0.030455999904208716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,64,0.003604444364706675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,256,0.01313333296113544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,32,0.0038888889054457345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,64,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,128,0.010441777606805166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,32,0.008020444048775567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,65536,0.22589066293504503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,65536,0.16445777151319715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,512,0.00868711123863856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,65536,0.3738844394683838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,51200,0.18636177645789254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,512,0.014535110857751636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,16384,0.05417066812515259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,65536,0.5251368946499294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,128,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,51200,0.1303537819120619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,51200,0.2940773434109158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,12288,0.04594044552909004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,16384,0.04428622126579285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,16384,0.09698222080866496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,10240,0.0362426671716902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,51200,0.41286489698621964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,12288,0.07281244463390775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,16384,0.14144710699717203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,8192,0.029455112086402044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,12288,0.10918044381671482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,10240,0.030564443932639226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,10240,0.06128799915313721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,7168,0.025807110799683466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,8192,0.023336889015303716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,8192,0.0502435564994812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,10240,0.09119911326302423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,6144,0.022039999564488728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,7168,0.021336000826623704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,8192,0.07522222068574694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,5120,0.01975200076897939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,6144,0.039184000757005476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,6144,0.05894577503204346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,5120,0.033356444703208074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,4096,0.015990222493807476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,12288,0.03418577710787455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,5120,0.051128000020980835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,4096,0.01513066722287072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,3584,0.014473777678277759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,4096,0.042471110820770264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,7168,0.044585777653588184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,3584,0.02552355494764116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,3072,0.012808889150619507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,3584,0.01402933398882548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,6144,0.01908266709910499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,7168,0.06583911180496216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,2560,0.01089333328935835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,3072,0.01292266696691513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,3584,0.03737066520584954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,2560,0.01975555552376641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,5120,0.016982222596804302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,2048,0.009471110999584198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,3072,0.03347999850908915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,2048,0.01092888911565145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,2560,0.029735998974906072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,4096,0.027727999620967444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,2048,0.016979555288950603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,1536,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,1024,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,2048,0.02566844390498267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,1536,0.010057777994208867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,1536,0.014135110709402295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,768,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,1024,0.011158222125636207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,128,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,1024,0.017603556315104168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,768,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,768,0.009004444711738164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,512,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,3072,0.022582223018010456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,256,0.0038293335172865125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,768,0.016019556257459853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,512,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,512,0.008344000412358178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,128,0.00318755561278926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,256,0.007383999725182851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,512,0.014455111490355598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,256,0.01261333293384976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,128,0.00812888890504837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,128,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,64,0.003702222059170405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,32,0.00388711111413108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,128,0.010151111417346531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,1024,0.009327111144860586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,65536,0.17752444744110107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,64,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,32,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,1536,0.021993777818149988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,51200,0.14612178007761636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,65536,0.1614346636666192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,65536,0.3280453417036268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,2560,0.012083555261294046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,16384,0.046875556310017906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,51200,0.2582373354170057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,51200,0.1286097764968872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,12288,0.03683200147416856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,16384,0.04362666606903076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,65536,0.38764267497592503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,51200,0.30414221021864146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,10240,0.029097778929604426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,16384,0.0858888891008165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,12288,0.03380977776315477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,12288,0.06480977932612102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,16384,0.10711822244856094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,8192,0.023591111103693645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,10240,0.029822223716311987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,10240,0.054794665839937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,12288,0.08180177874035306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,7168,0.020663999848895602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,8192,0.02316711180739933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,10240,0.06909244590335421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,6144,0.018616888258192275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,8192,0.0579377810160319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,7168,0.04023111197683547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,7168,0.05122399992412991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,6144,0.018597332967652213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,6144,0.03497777713669671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,5120,0.016100444727473788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,256,0.007991111112965478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,4096,0.013625777430004544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,5120,0.016763556334707472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,6144,0.04582666688495212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,5120,0.029959999852710303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,3584,0.01202755586968528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,4096,0.01494311127397749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,5120,0.0392231113380856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,4096,0.025124443901909724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,3072,0.010648888846238455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,3584,0.02288977801799774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,3584,0.013822221921549903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,2560,0.009127111070685917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,8192,0.04491466614935133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,3072,0.020185778538386028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,3072,0.013081777426931592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,3584,0.02951999836497837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,7168,0.020963556236690942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,2560,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,2560,0.017900443739361234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,3072,0.026776888304286536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,1536,0.006650666809744305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,2048,0.015536889433860779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,2560,0.02383377816942003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,1024,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,1536,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,1536,0.012982221941153208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,2048,0.021017778250906203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,768,0.004580444345871608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,1536,0.018223111828168232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,1024,0.01051644484202067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,768,0.009345778160625035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,768,0.008971555365456475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,1024,0.0155395550860299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,4096,0.03267644511328803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,768,0.013815110756291283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,512,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,2048,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,512,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,256,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,512,0.012452444268597497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,128,0.002985777747299936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,256,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,2048,0.01092977739042706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,64,0.0028311111446883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,128,0.008005333443482717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,32,0.0028568889117903183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,128,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,64,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,128,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,1024,0.009360888765917884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,65536,0.15215199523501927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,32,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,51200,0.11978577242957221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,512,0.0038746665749284956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,65536,0.2837831179300944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,51200,0.22296534644232857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,16384,0.03997866643799676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,65536,0.3815306557549371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,51200,0.3073759873708089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,16384,0.07430489195717706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,12288,0.0281315545241038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,16384,0.033033778270085655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,256,0.011377777490350934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,16384,0.10463288757536147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,10240,0.02297244469324748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,12288,0.02609866691960229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,12288,0.05686399671766493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,8192,0.01868711080816057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,10240,0.022488888767030504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,10240,0.048114667336146034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,12288,0.08110933171378241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,65536,0.11785954899258083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,7168,0.01631733278433482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,8192,0.017453332742055256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,8192,0.03962577713860406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,256,0.0032115555885765287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,51200,0.0950160026550293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,6144,0.014279999666743808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,8192,0.05667377842797173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,7168,0.015636444091796875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,5120,0.01257777793539895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,6144,0.014500444134076437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,7168,0.050833778248892895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,6144,0.031176000833511353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,4096,0.010330666270520952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,5120,0.026509332987997267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,5120,0.013152889079517789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,6144,0.04471910993258158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,3584,0.009568000005351173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,4096,0.011767110890812345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,5120,0.038134220573637224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,4096,0.02223377757602268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,3072,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,3584,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,3584,0.020336000455750358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,10240,0.06806221935484144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,3072,0.018273777431911893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,2560,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,3584,0.028926223516464233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,2560,0.010151111417346531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,3072,0.02616888946957058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,2560,0.016103999482260812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,2560,0.023431999815834895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,2048,0.013941332697868347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,2048,0.00960977789428499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,2048,0.020599999361568026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,1536,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,1536,0.011649777491887411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,4096,0.031832890378104314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,3072,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,1536,0.017874666386180453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,1024,0.009486222432719337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,1024,0.008328888979223039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,7168,0.03530222177505493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,1024,0.015237333046065437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,512,0.0034968890249729156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,768,0.008640888664457533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,768,0.008090666598743862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,512,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,256,0.0031271111220121384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,512,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,1536,0.005761777775155173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,256,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,512,0.012435555458068848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,128,0.00254133323000537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,256,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,1024,0.004577777865860197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,64,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,256,0.011071110765139261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,128,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,32,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,128,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,65536,0.08272711435953777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,768,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,64,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,128,0.0086666668454806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,32,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,51200,0.06491377618577746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,2048,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,16384,0.02473066747188568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,65536,0.11710577540927464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,65536,0.2422808806101481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,51200,0.0941306683752272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,51200,0.19134310881296793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,12288,0.01940444442960951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,16384,0.031910220781962075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,768,0.013635555903116861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,16384,0.06341155370076497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,10240,0.018961777289708454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,65536,0.3656755553351508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,12288,0.024473778075642053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,51200,0.2886906729804145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,8192,0.015969778100649517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,16384,0.10180799828635322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,10240,0.022107554806603327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,10240,0.04129777683152093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,7168,0.01425866617096795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,12288,0.07716355721155803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,8192,0.017044444878896076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,8192,0.034041778908835515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,10240,0.06569600105285645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,7168,0.01556799974706438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,7168,0.03052799900372823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,5120,0.010670222342014313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,6144,0.014327999618318347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,6144,0.026766222384240892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,7168,0.04907555712593926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,4096,0.009413333402739631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,6144,0.04235466652446323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,4096,0.019560888409614563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,5120,0.03570577833387587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,4096,0.011704000333944956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,3584,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,4096,0.030981332063674927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,6144,0.013210666676362356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,3584,0.017907554904619854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,3072,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,3584,0.011053333679835001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,8192,0.05505777729882134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,3584,0.028262221150928076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,2560,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,3072,0.010412444670995077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,3072,0.0159244438012441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,5120,0.023185777995321486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,2048,0.005031110925806893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,2560,0.009999111294746399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,3072,0.02512888941499922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,2048,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,2560,0.022679110368092854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,1536,0.004344888859324985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,12288,0.048713776800367564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,2048,0.019654222660594516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,1536,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,1024,0.0035048888789282907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,1536,0.010405333505736457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,1536,0.01720533271630605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,1024,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,1024,0.0088755554623074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,768,0.003208000005947219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,1024,0.014469333820872836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,512,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,768,0.007777777810891469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,5120,0.013028444515334235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,2560,0.01419911119672987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,768,0.01273155543539259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,256,0.0025031111306614345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,512,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,512,0.006724444528420766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,128,0.0023440000497632557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,2048,0.012416889270146688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,256,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,64,0.0021946666141351065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,128,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,128,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,256,0.010822222464614444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,65536,0.05880177683300442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,128,0.00831911133395301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,64,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,32,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,51200,0.04884799983766344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,65536,0.11673155095842148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,65536,0.20439555909898546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,51200,0.09310755464765762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,65536,0.3592391014099121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,256,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,512,0.01221333361334271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,51200,0.28619999355740017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,16384,0.054189334313074745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,12288,0.01205599970287747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,32,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,16384,0.031952887773513794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,10240,0.011037333144081963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,12288,0.0240328891409768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,16384,0.10059466626909043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,12288,0.04126311010784573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,8192,0.009760888914267222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,10240,0.03536533315976461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,768,0.008005333443482717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,12288,0.07648799816767375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,7168,0.009665778113736046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,8192,0.016976000534163583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,8192,0.028946667909622192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,10240,0.06501866711510552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,6144,0.009031111167536842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,51200,0.1607608927620782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,16384,0.015602666470739575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,7168,0.025877333349651758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,7168,0.01550222271018558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,5120,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,8192,0.053977777560551964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,6144,0.01425066590309143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,6144,0.02318488889270359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,5120,0.013075555364290873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,4096,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,7168,0.047375109460618764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,5120,0.020237333244747586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,3584,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,6144,0.04155644443300035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,4096,0.011540444360838996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,3072,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,5120,0.035659554931852556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,4096,0.017218665944205392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,3584,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,3584,0.01573777695496877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,2560,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,3072,0.010656000011497073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,4096,0.03056266572740343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,10240,0.022034666604465906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,2048,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,2560,0.010122666756312052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,2560,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,3072,0.025060445070266724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,1536,0.004006222304370669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,2048,0.011151110960377587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,2048,0.019945777124828763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,1536,0.009666666388511658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,1024,0.0034977777136696708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,1536,0.008920888933870528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,1536,0.016917333006858826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,1024,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,768,0.003154666680428717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,1024,0.008349333372380998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,1024,0.014484445254007975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,3072,0.014327999618318347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,3584,0.02749600013097127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,768,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,512,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,768,0.00796977761718962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,256,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,768,0.012501333322789935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,512,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,2048,0.00941422250535753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,512,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,128,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,256,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,512,0.011377777490350934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,64,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,128,0.007256000406212277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,256,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,32,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,65536,0.05706310934490628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,64,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,128,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,51200,0.046912001238928906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,32,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,16384,0.018565333551830716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,12288,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,65536,0.20441333452860513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,51200,0.16036266750759548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,10240,0.01015733347998725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,16384,0.05415022373199463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,8192,0.008629333641793992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,12288,0.041379554404152766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,6144,0.00999111102686988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,2560,0.022328888376553852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,10240,0.03460622164938185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,5120,0.008223111430803934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,256,0.007335999773608313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,8192,0.02864355511135525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,4096,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,7168,0.025813332862324182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,3584,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,3072,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,6144,0.022973333795865376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,2560,0.005107555538415909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,2048,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,128,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,5120,0.019951999187469482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,1536,0.003871110992299186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,4096,0.017120889491505094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,1024,0.0031866667171319327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,3584,0.015573332707087198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,3072,0.014206222361988492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,768,0.0028533333291610083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,512,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,2560,0.012671999633312225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,256,0.0022960000981887183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,2048,0.011085333095656501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,1536,0.009638222555319468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,1024,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,64,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,768,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,32,0.0021280000607172647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,256,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,128,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,51200,0.044545776314205594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,64,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,32,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,16384,0.012455110748608908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,12288,0.009932444327407414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,65536,0.2045902278688219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,10240,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,16384,0.054120000865724355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,51200,0.16037066777547201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,12288,0.04155288802252875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,7168,0.007704888780911763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,10240,0.034690668185551964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,8192,0.02863466739654541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,6144,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,5120,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,128,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,4096,0.008017777568764156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,512,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,3584,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,65536,0.057368887795342334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,5120,0.01995022263791826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,3072,0.0064702220261096954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,2560,0.005759999983840519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,7168,0.01071466671095954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,2048,0.004548444516128964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,3584,0.015797333584891427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,3072,0.014202665951516895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,1536,0.0038382220599386427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,2560,0.01274222218328052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,1024,0.003175111073586676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,2048,0.011155555645624796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,768,0.002874666617976295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,1536,0.009463999834325578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,512,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,768,0.007297777467303806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,7168,0.02574577762020959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,256,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,512,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,128,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,64,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,6144,0.022799111074871484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,32,0.0018586667461527719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,256,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,4096,0.017051556044154696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,128,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,64,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,32,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,8192,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,1024,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,16384,0.9802044762505425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,12288,2.429507573445638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,16384,3.0891884697808156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,12288,0.7257164319356283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,12288,2.4992408752441406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,16384,3.275835461086697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,10240,0.6157751083374023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,51200,3.5995886060926647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,8192,1.5855413013034398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,10240,2.0026141272650824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,8192,0.5602373547024196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,10240,1.9436995188395183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,8192,1.471326192220052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,7168,0.4715920024447971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,6144,0.4007359875573053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,6144,1.2023821936713326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,7168,1.3089510599772136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,7168,1.5789662467108831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,6144,1.1065990659925673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,5120,1.1272417704264324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,4096,0.7532835536532931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,5120,0.9277635150485568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,4096,0.27882933616638184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,3584,0.2498088942633735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,4096,0.7526444329155816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,3584,0.7131644354926215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,3072,0.5607395701938206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,3072,0.21574755509694418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,3584,0.6513128810458714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,5120,0.3303742143842909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,2560,0.5131155649820963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,2560,0.18948266241285536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,3072,0.5674337810940212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,2048,0.34976355234781903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,2560,0.4934488932291667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,2048,0.17557244830661348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,1536,0.29844265513949925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,1536,0.12769333521525064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,1024,0.22220088375939262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,2048,0.384335994720459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,768,0.18540710873074004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,1024,0.10365955697165595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,1536,0.31576087739732533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,768,0.08931377861234878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,512,0.136336882909139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,1024,0.256607108645969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,256,0.09064710802502102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,768,0.18517688910166422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,512,0.08130577537748548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,128,0.06446488698323567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,256,0.07389866643481784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,512,0.16206133365631104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,64,0.05811555518044365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,128,0.06960888703664143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,256,0.14589067300160727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,32,0.060023109118143715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,128,0.13322577211591932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,51200,10.075518290201822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,64,0.13691200150383845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,32,0.13521156046125624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,16384,2.2984301249186196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,51200,2.884680006239149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,51200,13.873942057291666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,65536,3.882804446750217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,16384,2.493306689792209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,16384,0.8861697514851888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,12288,1.7806943257649739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,12288,0.5869128968980577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,10240,0.4923573070102268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,10240,1.954483456081814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,12288,2.040210723876953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,51200,7.3561596340603295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,10240,1.5387253231472438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,8192,0.397167099846734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,8192,1.2739813062879775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,7168,1.03548706902398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,7168,0.3501946661207411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,7168,1.0162169138590496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,8192,1.3223608864678276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,65536,9.55978478325738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,6144,0.9657128651936849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,6144,0.30783555242750377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,6144,0.9062568876478406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,5120,0.8162284427218967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,5120,0.27976621521843803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,4096,0.619040913052029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,4096,0.2263608773549398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,4096,0.5676933394538032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,3584,0.5676853391859267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,5120,0.7211146884494357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,3584,0.1907191144095527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,3584,0.5302791065639919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,3072,0.1746951076719496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,3072,0.4764675564236111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,51200,10.849308437771269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,2560,0.417248010635376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,3072,0.45608891381157773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,2560,0.1536453300052219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,2048,0.3785155614217122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,2048,0.13819021648830837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,1536,0.2554950979020861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,2560,0.3722631136576335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,2048,0.3350959883795844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,1024,0.18868800004323324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,1536,0.11218044492933486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,1024,0.07959200276268853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,768,0.13363911045922175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,1536,0.23922223514980742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,512,0.10821244451734756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,768,0.0790168907907274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,512,0.06531911426120333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,1024,0.18591554959615073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,768,0.15137333340115017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,512,0.1277351114484999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,128,0.052779555320739746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,64,0.04673599865701464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,256,0.05851466788185967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,32,0.04885866575770908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,256,0.1137111054526435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,128,0.05637066894107395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,128,0.10633244117101033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,64,0.10708533393012153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,32,0.10697333017985027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,256,0.06691288948059082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,65536,14.129337734646269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,51200,0.8886204295688205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,65536,1.170556492275662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,16384,0.7162728839450412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,16384,0.3248915672302246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,12288,0.7798746426900228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,16384,0.8401261965433756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,12288,0.19684888256920707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,51200,2.550609800550673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,12288,0.5962435404459635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,10240,0.16953421963585746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,65536,3.149500528971354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,8192,0.4073475466834174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,10240,0.4882195790608724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,8192,0.13679467307196722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,8192,0.3835679954952664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,7168,0.33869067827860516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,51200,3.277210659450955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,7168,0.1207617786195543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,6144,0.296238210466173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,5120,0.233604457643297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,7168,0.4824470943874783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,6144,0.2877359920077854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,6144,0.10561422506968181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,4096,0.18790133794148764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,5120,0.23970754941304526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,4096,0.07457066906823052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,3584,0.16761155923207602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,3584,0.06680355469385783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,10240,0.4931902355617947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,3072,0.12907466623518202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,2560,0.11850844489203559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,3584,0.17167111237843832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,3072,0.05939022037718031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,3072,0.14895110660129124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,2048,0.11004355880949233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,2560,0.052871998813417226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,1536,0.07607377900017633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,2560,0.12653599845038518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,65536,4.979053497314453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,2048,0.045407109790378146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,1024,0.05108711123466492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,2048,0.10442222489251031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,768,0.04312710960706075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,1536,0.03706044289800856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,512,0.032695998748143516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,5120,0.09217955668767293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,1536,0.08371466398239136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,768,0.029105779197480943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,1024,0.0813573333952162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,512,0.026540444956885442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,256,0.02966844373279148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,128,0.018252443936136033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,768,0.051295999023649425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,256,0.03949600127008226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,4096,0.22325065400865343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,64,0.01585422290696038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,128,0.023403555154800415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,32,0.01644533375898997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,128,0.03747289048300849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,64,0.03744533326890733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,32,0.037646220790015325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,1024,0.03136000037193298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,256,0.022454222043355305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,512,0.04383466641108195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,65536,0.880425771077474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,16384,0.518630239698622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,51200,0.6687697834438748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,16384,0.19993333021799722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,12288,0.4084915584988064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,51200,1.9226712120903864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,12288,0.15006755457984075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,65536,2.4038382636176214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,16384,0.6469359927707249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,10240,0.35752532217237687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,12288,0.5146835645039877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,10240,0.12846488422817653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,8192,0.10787910885281032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,8192,0.29062043295966256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,7168,0.2640533447265625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,10240,0.3751173284318712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,8192,0.29503822326660156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,7168,0.2570426728990343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,6144,0.2228142155541314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,7168,0.0927377806769477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,65536,3.073828379313151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,6144,0.08073688877953424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,5120,0.19349333975050184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,51200,2.6606631808810763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,4096,0.15948622756534153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,6144,0.2404399977789985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,5120,0.07229244709014893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,3584,0.13483822345733643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,4096,0.05800355805291069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,3072,0.10655822356541951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,5120,0.18405511644151476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,3584,0.05225333240297106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,4096,0.15355111493004694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,3072,0.04986933204862806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,2048,0.07480533255471124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,3584,0.1420124504301283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,2560,0.04092622134420607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,1536,0.05961955255932278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,2048,0.03698577814631992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,1024,0.042575998438729175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,1536,0.028883556524912517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,768,0.03135111265712314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,1536,0.06445066796408759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,1024,0.024491555160946313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,512,0.023704888092146978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,1024,0.05168977710935804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,768,0.021567111214001972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,256,0.017415111263593037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,768,0.039897776312298246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,512,0.021134222547213238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,128,0.014681778020328946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,512,0.03257066673702664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,256,0.01816444430086348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,64,0.012081777883900536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,256,0.028972443607118394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,128,0.01822933389080895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,32,0.01221333361334271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,128,0.027969777584075928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,2560,0.0938533345858256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,64,0.027463111612531874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,32,0.027793778313530818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,3072,0.1160773303773668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,2560,0.1002346674601237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,2048,0.0846782194243537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,16384,0.4530613157484267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,65536,0.715229352315267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,51200,0.5686720212300619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,16384,0.4909075631035699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,12288,0.4008302158779568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,12288,0.13192799356248644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,51200,1.6784888373480902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,12288,0.468589358859592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,10240,0.3508604367574056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,65536,2.185826619466146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,10240,0.3033493359883626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,10240,0.13405688603719076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,51200,1.9775644938151042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,7168,0.18798311551411948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,8192,0.107387555970086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,8192,0.34116445647345645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,16384,0.16761155923207602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,6144,0.1621626615524292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,7168,0.07978400256898668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,7168,0.25161422623528373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,5120,0.15788533952501085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,6144,0.08461778031455146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,4096,0.12261333730485703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,65536,2.8366675906711154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,5120,0.059839109579722084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,6144,0.21560177538130018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,4096,0.05182222194141812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,3584,0.11396622657775879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,5120,0.15765066941579184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,3584,0.04587822159131368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,4096,0.12457866138882107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,3072,0.1018986635737949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,2560,0.07687911060121325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,3072,0.03935911258061727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,3584,0.11111999882592095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,2048,0.06095733245213827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,2560,0.03543377916018168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,3072,0.11397600173950195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,1536,0.047489778863059155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,8192,0.2394719918568929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,2560,0.08357510964075725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,2048,0.030247999562157526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,768,0.027317333552572463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,1536,0.029083556599087183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,2048,0.06897333595487806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,1024,0.020890666378868952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,512,0.027126222848892212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,1024,0.042381332980261914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,768,0.01976799964904785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,256,0.020943111843532983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,512,0.017251556118329365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,768,0.03465955456097921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,128,0.012499555945396423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,512,0.028140445550282795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,256,0.016004444824324716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,64,0.010882666541470421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,32,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,128,0.015912888778580558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,128,0.023748444186316595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,64,0.023788443870014612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,32,0.023702222439977858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,1024,0.03480444351832072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,1536,0.05466755562358432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,65536,0.6052213244967991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,256,0.025755556093321905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,16384,0.39852266841464573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,51200,0.47846221923828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,65536,1.4985644022623699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,16384,0.14027289549509683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,12288,0.2964595423804389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,51200,1.3797182506985133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,16384,0.4422542254130046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,12288,0.33250223265753853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,10240,0.28624267048305935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,10240,0.115100449985928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,51200,1.6642160415649414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,8192,0.0855413344171312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,7168,0.2373253239525689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,65536,2.003152847290039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,10240,0.3244817786746555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,8192,0.2048471106423272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,6144,0.16426666577657065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,7168,0.07731199926800199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,5120,0.13489777512020534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,7168,0.1867226627137926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,6144,0.05786222219467163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,4096,0.10625688897238837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,5120,0.050196445650524564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,3584,0.09784622324837579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,5120,0.13765154944525823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,6144,0.15304799874623617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,4096,0.10461333062913682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,4096,0.04115822248988681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,3072,0.0795724458164639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,3584,0.037408000893063016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,3584,0.09338488843705918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,2560,0.06483644247055054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,3072,0.0331911113527086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,2048,0.05128355489836799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,12288,0.10811822281943427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,2560,0.030576888057920668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,2560,0.07078755564159818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,2048,0.025620443953408137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,8192,0.20656000243292916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,1024,0.029079112741682265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,1536,0.021391110287772283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,3072,0.07991911305321588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,768,0.021145777569876775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,512,0.016343111793200176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,1024,0.01796711153454251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,2048,0.057130667898390025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,1024,0.03533600105179681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,768,0.016582222448454965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,512,0.01514933341079288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,768,0.029150221082899306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,128,0.009679999616410997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,512,0.02330400049686432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,64,0.008794666992293464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,256,0.014084445105658637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,128,0.014281777871979607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,256,0.02069155540731218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,32,0.009126221968068017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,128,0.01994755533006456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,64,0.019899555378490023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,32,0.0199955552816391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,1536,0.03911822372012668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,1536,0.04572088850869072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,256,0.012885332935386233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,65536,0.5301102002461752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,65536,1.2642293506198459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,51200,0.9342675738864474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,16384,0.34699556562635636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,51200,0.4189093377855089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,12288,0.25174933009677464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,16384,0.12958755758073595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,16384,0.41076355510287815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,12288,0.11208800474802653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,10240,0.20229332976871064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,12288,0.2739208804236518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,10240,0.081423110432095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,8192,0.17087554931640625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,51200,1.4499724706013997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,7168,0.15887378321753606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,10240,0.23241776890224883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,6144,0.12967911031511095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,7168,0.06355555852254231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,8192,0.1884533299340142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,7168,0.16554666890038386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,5120,0.11069066656960382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,6144,0.1377448903189765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,4096,0.08409777614805435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,5120,0.04402133491304186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,3584,0.07960533433490329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,5120,0.1157733334435357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,4096,0.03613422314325968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,3072,0.06287733051511976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,4096,0.09183377689785427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,3584,0.033857779370413885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,2560,0.053467555178536304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,3584,0.08263555500242445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,3072,0.030210667186313208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,2048,0.04344177908367581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,3072,0.07119733095169067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,2560,0.02721244427892897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,1536,0.032933334509531655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,2560,0.06321599748399523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,2048,0.02383288906680213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,8192,0.06511733267042372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,2048,0.05195022291607327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,1536,0.0199955552816391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,768,0.019344889455371432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,1536,0.041562666495641075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,1024,0.016875555117925007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,512,0.013977777626779346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,768,0.015201777219772339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,1024,0.031742221779293485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,256,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,512,0.014145778285132514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,128,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,512,0.021221333079867895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,256,0.013312000367376538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,64,0.007970666719807519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,256,0.018889778190188937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,32,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,128,0.013291555146376291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,128,0.0180711117055681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,64,0.018233777748213876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,32,0.0181377778450648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,65536,1.7717431386311848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,1024,0.02316444449954563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,6144,0.05156533254517449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,65536,0.4803502294752333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,768,0.026278222600619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,51200,0.3786915673149957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,51200,0.842282666100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,16384,0.11098666985829671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,16384,0.30281511942545575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,12288,0.22951110204060873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,10240,0.19186755021413168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,12288,0.08437689145406087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,16384,0.39644267823961044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,51200,1.124432881673177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,8192,0.15598666667938232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,65536,1.5035137600368922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,12288,0.27802666028340656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,10240,0.07708622349633111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,10240,0.21240622467464873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,6144,0.11714666419559056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,8192,0.05877244472503662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,7168,0.05245155427190992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,8192,0.1746631066004435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,5120,0.10033955838945176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,6144,0.047338667843076915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,4096,0.07520711421966553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,5120,0.03916622201601664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,6144,0.12172088358137344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,3584,0.05903377797868517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,5120,0.10118133491939968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,4096,0.03344088792800903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,65536,1.1325129403008354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,3072,0.058232890235053174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,4096,0.08278666602240668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,3584,0.030044443077511255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,2560,0.043262223402659096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,3584,0.07323200172848172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,3072,0.027161777019500732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,2048,0.03936266567971971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,3072,0.06275377670923869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,2560,0.02456444501876831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,1536,0.030103110604816016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,2560,0.05446488989724053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,1024,0.020929776959949072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,2048,0.0214346663819419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,2048,0.04554577668507894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,1536,0.017870222528775532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,768,0.016334222422705758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,1024,0.015285332997639975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,7168,0.13581511709425184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,512,0.012880889077981314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,768,0.014450665977266101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,256,0.009480888644854227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,768,0.023810666468408372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,512,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,7168,0.15403110451168484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,128,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,512,0.0188400000333786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,64,0.0069164443347189165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,256,0.012258666257063547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,256,0.016866667403115165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,128,0.01623644431432088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,64,0.016453334026866488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,32,0.016191111670600045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,1536,0.036791112687852644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,1024,0.027679110566775005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,65536,0.42542844348483616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,65536,1.008968883090549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,128,0.012473777764373355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,51200,0.33412978384229874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,16384,0.278165340423584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,51200,0.7853520181443957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,65536,1.3384257422553167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,12288,0.2165271176232232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,16384,0.26684178246392143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,16384,0.1009991102748447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,12288,0.07844711012310453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,10240,0.1522293355729845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,10240,0.06896177927652995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,51200,1.0031858020358615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,12288,0.2047679954104953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,10240,0.1663866705364651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,7168,0.11672000090281169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,8192,0.05215911070505778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,6144,0.10255199670791626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,8192,0.1388746632470025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,32,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,5120,0.08347821897930569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,6144,0.04702933298216926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,6144,0.10498933659659491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,5120,0.03479111194610596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,5120,0.08646133210923936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,3584,0.056941330432891846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,4096,0.07315911187065972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,3072,0.04825777808825175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,3584,0.026833777626355488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,3584,0.06213066975275675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,2560,0.04186044467820061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,3072,0.024596444434589807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,3072,0.0544177790482839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,8192,0.13562933603922525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,2048,0.03210399879349603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,2560,0.022287110487620037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,2560,0.04735733403099907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,1536,0.02557777696185642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,2048,0.01905066602759891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,7168,0.04630577895376417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,1024,0.018389332625601027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,2048,0.03955555624432034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,768,0.01496088835928175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,7168,0.1194595495859782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,1024,0.014274666706720987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,4096,0.06838133600023058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,512,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,768,0.013206221991115145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,1024,0.02458400030930837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,256,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,768,0.02093066606256697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,4096,0.029829333225886028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,512,0.012082666986518435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,128,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,512,0.016819554898473952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,64,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,256,0.011546666423479715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,128,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,32,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,256,0.015186667442321777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,128,0.014450665977266101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,64,0.014499555031458536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,32,0.014488000008794995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,65536,0.36050576633877224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,1536,0.016509332590632968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,51200,0.2834862338172065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,51200,0.6597697999742296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,16384,0.22664266162448457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,65536,0.9700159496731229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,12288,0.17471910847557914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,16384,0.08390755785836114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,10240,0.1365164385901557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,65536,1.1142106586032443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,51200,0.795165326860216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,12288,0.07480000125037299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,8192,0.10841333203845555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,12288,0.22743821144104004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,10240,0.06341866652170818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,7168,0.10505244466993545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,10240,0.1374853319591946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,8192,0.048135999176237315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,8192,0.1297066609064738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,6144,0.08243733644485474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,7168,0.05311110946867201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,5120,0.06703466839260526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,7168,0.09884622361924912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,6144,0.03578666514820523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,4096,0.04981422093179491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,6144,0.08857778045866226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,5120,0.030985779232449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,3584,0.04636444316969978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,1536,0.03180266751183404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,4096,0.026166222161716882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,5120,0.1000666684574551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,3072,0.037090665764278836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,4096,0.05835022528966268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,2560,0.031239112218221027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,3584,0.023797333240509033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,3584,0.05261422197024027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,3072,0.02162133322821723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,16384,0.21642843882242838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,3072,0.04638044370545281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,1536,0.020960888928837244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,2048,0.01687466601530711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,1024,0.016155555844306946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,2048,0.03439377744992574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,1536,0.014319111903508505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,768,0.013252444565296173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,1536,0.02702577743265364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,1024,0.012514666550689273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,512,0.010327111515733931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,1024,0.020933333370420668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,768,0.011782222323947482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,256,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,768,0.01811377704143524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,512,0.01457155578666263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,128,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,256,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,256,0.01309333327743742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,64,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,128,0.01052266690466139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,32,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,2048,0.027161777019500732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,2560,0.01927199959754944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,64,0.01240711079703437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,2560,0.03996088769700792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,32,0.012784889174832238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,65536,0.5506381988525391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,65536,0.2945600085788303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,51200,0.23512000507778594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,51200,0.5317084524366591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,16384,0.15858844916025797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,512,0.011143111520343356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,12288,0.14297955565982395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,16384,0.07138222455978394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,12288,0.05596888727611966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,10240,0.11810489495595296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,16384,0.19738045003679064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,65536,0.916776869032118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,51200,0.7226195335388184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,8192,0.09577511416541205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,12288,0.1517137818866306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,10240,0.04867911007669237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,7168,0.08203466733296712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,128,0.012392000191741519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,8192,0.03838488790724013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,10240,0.12483289506700303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,6144,0.07086310784022014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,5120,0.05962844689687093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,8192,0.10082044204076131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,6144,0.03052088949415419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,4096,0.04730133215586344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,5120,0.02683199942111969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,7168,0.09701510932710435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,3584,0.042043556769688926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,6144,0.07731822464201185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,3072,0.03584533267550998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,5120,0.06578755378723145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,4096,0.02314400010638767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,2560,0.02845066785812378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,3584,0.022244445151752893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,4096,0.053038222922219165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,3072,0.01924000018172794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,2048,0.02384088933467865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,3584,0.0473671125041114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,1536,0.018940443793932598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,2560,0.017323555217848886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,3072,0.0419680012596978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,1024,0.014090667168299357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,2048,0.015212444795502557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,2560,0.03644355469279819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,1536,0.014116444521480136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,768,0.011183111204041375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,2048,0.031086222993002996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,1536,0.02442399991883172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,1024,0.0120328888297081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,768,0.011134222149848938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,1024,0.01921244462331136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,256,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,768,0.016580444243219163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,512,0.010413332945770687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,128,0.005758222192525864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,256,0.009895111123720804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,64,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,256,0.012150222228633033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,7168,0.034396443102094866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,128,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,32,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,128,0.011687111523416309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,64,0.011640888949235281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,32,0.011711999773979187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,65536,0.28025155597262913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,65536,0.5914337899949815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,51200,0.3905324406094021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,512,0.0086666668454806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,16384,0.16161599424150255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,51200,0.21985244750976562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,12288,0.1228675577375624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,16384,0.067413330078125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,65536,0.7945875591701932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,512,0.013430222041077085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,10240,0.09362222088707818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,16384,0.17768800258636475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,51200,0.6178373230828179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,12288,0.06545155578189425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,8192,0.07508889171812269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,10240,0.04605688982539707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,7168,0.07138044304317899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,12288,0.13321955998738608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,8192,0.03626222080654568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,10240,0.11201600233713786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,6144,0.060497780640920006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,5120,0.04752799868583679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,8192,0.09000088771184285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,6144,0.02909244431389703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,7168,0.0789511137538486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,5120,0.026390223039521113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,5120,0.06291644440756904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,3584,0.03380444314744737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,4096,0.022086222966512043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,3072,0.02829244401719835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,4096,0.047487109899520874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,3584,0.020385776956876118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,2560,0.024655999408827886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,3072,0.01847644481394026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,3072,0.03804711169666714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,2048,0.020050666398472257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,2560,0.01682577696111467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,2560,0.03269955515861511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,1536,0.016344888342751395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,2048,0.015029332704014249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,1024,0.012496888637542725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,2048,0.02755466600259145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,1536,0.01348888874053955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,768,0.010545777777830759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,7168,0.03285955720477634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,1536,0.022310222188631695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,1024,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,512,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,1024,0.0174319992462794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,256,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,768,0.010960000256697336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,6144,0.06938311126497057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,768,0.015202666322390238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,128,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,512,0.009992000129487779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,64,0.004531555705600315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,512,0.012425777812798819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,256,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,32,0.00479288895924886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,128,0.009677333136399588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,128,0.01075111081202825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,3584,0.04316088888380262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,64,0.010677333507272931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,32,0.010675555302037133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,65536,0.45268090565999347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,65536,0.25665420956081814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,51200,0.18962221675448945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,16384,0.12812533643510607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,12288,0.09490755531522964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,16384,0.05817955732345581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,65536,0.6738088925679525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,51200,0.5235573450724283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,16384,0.15321955415937635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,12288,0.046514666742748685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,10240,0.09450222386254205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,8192,0.0767422252231174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,12288,0.11865688694847955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,10240,0.04469511244032118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,256,0.011085333095656501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,7168,0.06654044654634264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,10240,0.09598310788472493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,8192,0.031833777825037636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,8192,0.07823555337058173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,6144,0.05508888761202494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,7168,0.03183822168244256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,5120,0.04154400030771891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,6144,0.0274008893304401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,7168,0.06936888562308417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,6144,0.060544888178507485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,4096,0.03507377703984579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,5120,0.02426044477356805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,3584,0.03113599949412876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,51200,0.40374310811360675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,4096,0.019891555110613506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,3072,0.02737244466940562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,4096,0.04180888997183906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,3584,0.018020444446139865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,2560,0.022967999180157978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,3072,0.016321778297424316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,4096,0.040030221144358315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,2048,0.019655111763212416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,3072,0.033259554041756526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,2560,0.015230221880806817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,1536,0.015395555231306287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,2560,0.028860443168216284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,1024,0.011869333684444427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,2048,0.024451555477248296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,1536,0.020044444335831534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,1024,0.010451555252075195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,768,0.009728888670603434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,1024,0.015504888362354703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,512,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,768,0.009999111294746399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,768,0.013761777844693927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,256,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,512,0.009376889301670922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,5120,0.05122577812936571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,128,0.004584889031118817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,256,0.00906399968597624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,3584,0.03762133253945245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,256,0.0103493332862854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,64,0.00453333349691497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,128,0.009072889056470659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,32,0.004847110973464118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,128,0.009724444813198512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,2048,0.014543111125628153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,64,0.009733333355850643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,1536,0.012044444680213928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,32,0.009730666875839233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,65536,0.3620764414469401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,16384,0.09452355570263332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,51200,0.33790132734510636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,51200,0.16815288861592612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,12288,0.08624977535671657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,16384,0.056418663925594754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,65536,0.5265973409016927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,10240,0.06313155757056342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,16384,0.1318302286995782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,51200,0.40975554784138996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,512,0.011204444699817233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,12288,0.04892444279458788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,8192,0.048601776361465454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,12288,0.09803466664420234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,7168,0.04715911216206021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,10240,0.03752888904677497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,6144,0.04142489035924276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,8192,0.02985244327121311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,10240,0.08271733257505628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,7168,0.02687377731005351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,5120,0.032827556133270264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,8192,0.06830310821533203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,7168,0.0598542226685418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,4096,0.02696977721320258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,6144,0.02472266720400916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,3584,0.024433778391944036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,5120,0.020971554848882887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,6144,0.05206222335497538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,3072,0.021143111917707656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,4096,0.018068444397714403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,2560,0.01741155485312144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,4096,0.036426666710111834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,3584,0.032842665910720825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,3072,0.015464888678656684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,2048,0.015249777171346875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,65536,0.2111582226223416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,3072,0.028928889168633357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,1536,0.012420444852775998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,2560,0.014211555322011312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,1024,0.009313777916961247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,2048,0.013164444102181328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,2560,0.024749333659807842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,768,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,1536,0.011731555892361535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,2048,0.02146044373512268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,512,0.006421333385838403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,1024,0.010201777848932479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,1536,0.01738933391041226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,1024,0.013761777844693927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,256,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,768,0.012106666962305704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,512,0.009935999910036722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,128,0.004182222402758068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,256,0.008611555728647444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,256,0.009107555780145857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,64,0.0038559999730851916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,128,0.008431110945012834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,32,0.0042133331298828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,128,0.008804444637563493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,5120,0.04369599951638115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,3584,0.01684266659948561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,64,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,32,0.008833777573373582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,65536,0.3453048865000407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,65536,0.16477778222825792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,51200,0.2177679936091105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,16384,0.08787377675374348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,51200,0.13158844576941595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,65536,0.4189324378967285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,12288,0.0652079979578654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,512,0.009083555804358589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,16384,0.044758223825030856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,51200,0.3318462106916639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,16384,0.10982666412989299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,10240,0.054914666546715625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,8192,0.04047288828425937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,10240,0.030776888132095337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,12288,0.08076533344056872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,7168,0.03672622309790717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,8192,0.02440711193614536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,10240,0.06828266382217407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,6144,0.030977778964572485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,8192,0.055604444609748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,6144,0.019959110352728102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,7168,0.04942044284608629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,4096,0.021359999974568684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,6144,0.043113778034845986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,5120,0.017840888765123155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,5120,0.03675288955370585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,3584,0.019620445039537217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,4096,0.015415110521846347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,3072,0.017567111386193168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,3584,0.014490667316648694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,3584,0.0273973329199685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,768,0.009712888962692684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,2560,0.015077332655588785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,3072,0.01346666696998808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,3072,0.024452444579866197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,2560,0.012139555480745105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,12288,0.03640088770124648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,2560,0.02107288936773936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,2048,0.011077333655622272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,2048,0.018262222409248352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,1024,0.007756444315115611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,7168,0.022259554929203455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,1536,0.010216889282067617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,1536,0.0151235560576121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,768,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,1024,0.009380444056457942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,1024,0.012071111136012606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,512,0.005708444449636672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,768,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,256,0.004559999952713649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,768,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,512,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,128,0.003859555555714501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,512,0.008963555925422246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,256,0.008054222497675154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,64,0.0032462223122517266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,2048,0.012710221939616732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,128,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,32,0.003834666477309333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,1536,0.010268444816271463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,64,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,65536,0.23697156376308867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,32,0.007768889268239339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,5120,0.02732177740997738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,51200,0.18040266301896837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,16384,0.06143289142184787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,65536,0.32628001107109916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,51200,0.2562471230824788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,12288,0.044263998667399086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,16384,0.04201155569818285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,256,0.008049777812427944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,16384,0.09032977951897515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,10240,0.03756355577045017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,12288,0.03295822276009454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,128,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,8192,0.029425776667065088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,12288,0.06391911374198066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,7168,0.025123554799291823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,10240,0.05396711164050632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,8192,0.022580444812774658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,6144,0.023001778456899855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,8192,0.04425599839952257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,7168,0.020548444655206468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,5120,0.019375110665957134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,7168,0.039721777041753135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,6144,0.01830933325820499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,65536,0.15754932827419704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,4096,0.014881778094503613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,5120,0.016563556260532804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,5120,0.029822223716311987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,51200,0.12520533137851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,3584,0.013691555294725629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,4096,0.014450665977266101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,3072,0.012084444363911947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,4096,0.02438666588730282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,3584,0.013794666363133324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,4096,0.031066666046778362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,2560,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,3584,0.022266666094462078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,3072,0.012543111211723752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,2048,0.008967999782827165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,2560,0.011509333219793109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,1536,0.007530666887760162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,10240,0.029278222057554457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,2560,0.017516444126764934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,2048,0.010337777435779572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,1024,0.005966222120655908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,2048,0.0151555554734336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,768,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,1536,0.01276622215906779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,1024,0.008976000050703684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,512,0.004256000121434529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,1024,0.010456889039940305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,768,0.008575110799736446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,256,0.0035537779331207275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,6144,0.034585777256223894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,768,0.009078222016493479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,512,0.008005333443482717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,128,0.0031902222997612427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,512,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,256,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,64,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,32,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,128,0.007678222325113084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,256,0.007270221908887227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,128,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,64,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,32,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,65536,0.154874669180976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,51200,0.14714044994778103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,3072,0.020049777295854356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,16384,0.04634133312437269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,65536,0.2793324523501926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,12288,0.03297066688537598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,1536,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,51200,0.12337599860297309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,16384,0.04161422120200263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,10240,0.027612444427278306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,16384,0.0728915532430013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,12288,0.03221688999070062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,8192,0.02259911100069682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,10240,0.028897778855429754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,7168,0.019811555743217468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,10240,0.047248890002568565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,8192,0.022408000297016565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,6144,0.017312889297803242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,8192,0.03874933388498094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,7168,0.020272889071040683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,5120,0.014568888478808932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,7168,0.03479111194610596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,6144,0.018204443984561496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,6144,0.030680888228946265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,65536,0.1841226683722602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,4096,0.012435555458068848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,3584,0.011232000258233813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,5120,0.016418667303191293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,5120,0.02644177774588267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,3072,0.010274666878912184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,3584,0.01346933344999949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,4096,0.022055110997623865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,2560,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,3584,0.019871110717455547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,3072,0.01240888900227017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,3072,0.017860444055663217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,2048,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,51200,0.22067288557688394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,2560,0.011471110913488599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,1536,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,1024,0.005496888938877318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,2560,0.015819556183285184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,12288,0.05531733234723409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,2048,0.01038933296998342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,768,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,1536,0.009679111341635386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,2048,0.013766222529941134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,512,0.0038648889296584656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,1024,0.008989333278603023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,256,0.0031848889258172777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,1536,0.011446221835083432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,768,0.008467555873923832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,1024,0.009548444714811113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,512,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,128,0.002871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,768,0.008343111309740279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,64,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,256,0.007631111476156447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,512,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,32,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,128,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,256,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,128,0.006310222049554189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,65536,0.13795199659135607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,64,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,32,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,4096,0.014473777678277759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,65536,0.11204711596171062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,16384,0.03642400105794271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,12288,0.02665422194533878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,51200,0.09047822157541911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,65536,0.23495732413397896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,16384,0.030449777841567993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,10240,0.021908443835046556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,51200,0.18561066521538627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,16384,0.0613937775293986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,12288,0.023775999744733173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,8192,0.017924444542990792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,7168,0.015981333123313058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,12288,0.04732444551255968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,10240,0.02174844510025448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,6144,0.0144213338692983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,8192,0.01663733356528812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,10240,0.0403164459599389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,7168,0.015511999527613321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,5120,0.012322666744391123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,8192,0.0328764451874627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,4096,0.010656000011497073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,6144,0.014120888378885059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,7168,0.029790222644805908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,6144,0.026504000027974445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,5120,0.012856889102194043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,4096,0.011431111229790581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,5120,0.022671999202834234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,3072,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,3584,0.011032889286677042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,4096,0.019055111540688407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,2560,0.007040888898902469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,3584,0.017298666967286002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,3072,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,51200,0.09919022189246283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,3072,0.015857777661747403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,2560,0.009992000129487779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,1536,0.005133333305517833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,2560,0.013845333622561561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,2048,0.009378666679064432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,1536,0.00869066682126787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,2048,0.012335999972290464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,768,0.0035617777870761026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,1536,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,1024,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,512,0.0032115555885765287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,1024,0.008868444297048781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,768,0.007956444389290279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,256,0.0028239999794297745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,768,0.007687999970383114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,512,0.00738933351304796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,128,0.0025315555847353404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,512,0.00666133314371109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,3584,0.008999110923873054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,256,0.0069857777820693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,64,0.002524444419476721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,32,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,256,0.006002666635645761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,128,0.007245333658324347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,65536,0.07351377937528822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,128,0.005691555639108022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,51200,0.059149331516689725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,64,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,2048,0.006222222414281633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,32,0.005655999812814925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,16384,0.020764445265134174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,12288,0.016757332616382174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,65536,0.18929689460330537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,51200,0.08967288997438218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,16384,0.029606223106384277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,10240,0.01566844516330295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,51200,0.14899555842081705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,16384,0.05031555559900072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,8192,0.012896888785892062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,12288,0.03882844580544366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,10240,0.02147377696302202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,7168,0.011454222102959951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,8192,0.016740444633695815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,10240,0.03328444560368856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,6144,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,8192,0.02773333258099026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,7168,0.01533600025706821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,5120,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,6144,0.014144000079896716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,4096,0.007520889242490132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,6144,0.021996445126003687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,5120,0.01926666663752662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,3584,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,4096,0.011400889191362592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,4096,0.016480889585283067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,3072,0.006078222145636876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,65536,0.11084266503651936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,3584,0.015170666906568738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,1024,0.004010666575696734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,2560,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,3072,0.010327111515733931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,3072,0.013752000199423896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,2048,0.004769777672158347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,2560,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,2560,0.01219733307758967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,1536,0.0041786668201287585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,2048,0.009302222066455418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,2048,0.010834666589895884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,1024,0.0035182221068276297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,7168,0.02482755482196808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,1536,0.008625778059164682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,768,0.0031591111587153543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,1024,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,1024,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,512,0.0028284444577164124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,768,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,768,0.007255111303594377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,3584,0.010804444551467896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,256,0.002467555511328909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,512,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,128,0.0021991110924217436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,512,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,256,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,64,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,256,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,12288,0.02343822187847561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,32,0.0021608887861172357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,128,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,65536,0.05996533234914144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,51200,0.04689777890841166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,32,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,65536,0.11089689201778835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,16384,0.016472889317406546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,65536,0.18937955962287056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,5120,0.012824888858530255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,51200,0.08948889043596055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,1536,0.009378666679064432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,51200,0.1484702163272434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,12288,0.012792000340090858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,16384,0.028730667299694482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,16384,0.05013422171274821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,8192,0.012102222277058495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,12288,0.03830755419201321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,10240,0.021335111724005804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,7168,0.010791111323568555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,10240,0.03268977668550279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,8192,0.016719111137919955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,6144,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,8192,0.027075555589463975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,7168,0.015238222148683337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,128,0.007199111084143321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,5120,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,7168,0.024295111497243244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,6144,0.013973333769374423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,4096,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,5120,0.012809777425395118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,6144,0.02167822255028619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,3584,0.0063031112982167145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,5120,0.018945778409639995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,4096,0.011429333024554782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,3072,0.005695999910434087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,3584,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,2560,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,3072,0.01035733355416192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,3584,0.014889778362380134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,2048,0.00453244439429707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,3072,0.013634666800498962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,10240,0.011511111425028907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,2560,0.009676444033781687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,12288,0.023120888405376013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,1536,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,2560,0.0120666664507654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,1024,0.00319377767542998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,2048,0.009331555830107795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,768,0.0028826666788922418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,2048,0.010751999914646149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,1536,0.008660444782839881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,1024,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,512,0.0025048889219760895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,1536,0.009312888814343346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,256,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,1024,0.007763555480374231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,512,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,128,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,256,0.007234666910436418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,512,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,64,0.0018791111393107309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,256,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,64,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,32,0.0021280000607172647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,128,0.007000000112586551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,128,0.0052933332820733385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,65536,0.054382221566306226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,51200,0.04263022210862902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,64,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,16384,0.011622221933470832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,32,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,12288,0.012143111063374413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,65536,0.1873333321677314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,51200,0.14797511365678576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,10240,0.01072177787621816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,16384,0.050256000624762646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,12288,0.03799644443723891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,7168,0.00849955528974533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,10240,0.03249511122703552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,768,0.007681777907742395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,8192,0.026862222287389968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,768,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,7168,0.024124445186720953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,4096,0.0069102222720781965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,6144,0.021615111165576514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,3584,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,5120,0.01888533267709944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,3072,0.005719111197524601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,4096,0.016192000773217943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,2560,0.005541333307822545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,3584,0.014783110883500842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,2048,0.004206222378545337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,3072,0.013595555391576556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,1536,0.003850666599141227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,8192,0.009073778159088558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,2560,0.012094222009181976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,1024,0.0031946665710873077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,768,0.0028719999310043124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,2048,0.01074399964676963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,512,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,1536,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,256,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,6144,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,1024,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,128,0.002143111079931259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,768,0.007236444287829929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,5120,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,32,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,512,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,256,0.005460444423887465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,65536,0.0480693346924252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,128,0.005067555440796746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,51200,0.039705776506000094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,64,0.004978666702906291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,16384,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,32,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,12288,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,10240,0.008309333688682979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,8192,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,4096,0.016230222251680162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,7168,0.006991110742092133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,65536,0.18885955545637342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,16384,0.049984001451068454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,51200,0.14789777331882054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,12288,0.03810844487614102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,10240,0.03226222263442145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,6144,0.010006222460005019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,5120,0.017708443933063082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,4096,0.008351999852392409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,3584,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,8192,0.026725333597924974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,3072,0.007207999626795451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,2560,0.009672889278994666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,7168,0.024346666203604803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,2048,0.004443555656406614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,6144,0.02157955533928341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,5120,0.018900444110234577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,1536,0.0038817777401871155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,4096,0.016203555795881484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,3584,0.014781332678265043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,1024,0.0032026666320032547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,3072,0.013759111364682516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,2560,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,512,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,256,0.0021964444054497615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,128,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,1536,0.008997333546479544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,1024,0.008015111088752747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,768,0.0070426662762959795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,64,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,32,0.0018791111393107309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,512,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,256,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,128,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,32,0.004967999955018361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,768,0.0028488888508743713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,2048,0.010739555789364709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,64,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,16384,0.8769955635070801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,16384,2.9733592139350047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,16384,3.212051603529188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,51200,3.4922907087537975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,12288,2.209747526380751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,16384,2.748679055107964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,12288,2.25855467054579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,12288,0.6822177569071451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,10240,0.5839493539598253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,10240,2.0722177293565536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,12288,2.0857813093397355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,10240,1.965293248494466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,10240,1.7528293397691515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,51200,8.963963826497396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,8192,2.011407216389974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,8192,1.4739333258734808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,8192,0.45769511328803164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,7168,0.4100995593600803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,8192,1.3597013685438368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,7168,1.4333502451578777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,7168,1.301605330573188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,6144,0.3656746546427409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,7168,1.1875698301527235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,6144,1.1374328401353624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,6144,1.4543742073906794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,5120,0.9224924511379666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,5120,0.3098337650299072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,6144,1.0078516006469727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,5120,0.9517093234592013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,4096,0.9380968941582574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,4096,0.25231022304958767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,4096,0.7428773244222006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,5120,0.8278506596883138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,3584,0.614832878112793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,4096,0.6458097563849555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,51200,13.291390313042534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,3584,0.23397776815626356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,3584,0.6585431098937988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,3072,0.5411866505940756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,3072,0.20383021566602919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,3584,0.5725590917799208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,2560,0.4484088685777452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,3072,0.5851253403557671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,2560,0.17912799782223174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,2048,0.3446711169348822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,3072,0.5022995736863878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,2048,0.15245778030819362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,1536,0.29913777775234646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,2560,0.45597510867648655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,51200,9.23062981499566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,2048,0.5182293256123861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,2048,0.3571182092030843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,1024,0.2005164490805732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,1536,0.307241784201728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,768,0.1464124388164944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,1024,0.09656711419423421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,1024,0.2446026537153456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,1536,0.28045688735114205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,1024,0.21419910589853922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,768,0.08501599894629584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,512,0.12402666939629449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,2560,0.46596442328559023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,256,0.08656711048550075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,512,0.0759484436776903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,768,0.17761333783467612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,512,0.15609867042965359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,128,0.06088799900478787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,256,0.1363875601026747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,512,0.13908799489339194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,64,0.057962669266594775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,256,0.1128142277399699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,128,0.06499822272194757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,32,0.05560355716281467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,128,0.13056888845231798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,128,0.08356089062160915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,64,0.13104711638556585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,32,0.13217688931359187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,1536,0.12044621838463677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,768,0.18307644791073271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,256,0.06813155280219184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,51200,2.64955817328559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,65536,3.3560061984592013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,51200,7.010300530327691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,16384,2.7560585869683156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,16384,2.784876505533854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,16384,0.696982224782308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,65536,9.083291795518663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,12288,1.814123577541775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,12288,0.5608204205830892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,12288,1.891832987467448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,51200,10.687278747558594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,16384,2.1365040673149958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,51200,7.1973317464192705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,10240,0.44402045673794216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,12288,1.5414888593885634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,10240,1.6400950749715169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,10240,1.4798355102539062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,8192,1.1812008751763237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,8192,0.42981688181559247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,65536,13.157920837402344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,10240,1.3112249374389648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,8192,1.2089404000176323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,7168,0.965362654791938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,7168,0.3252257770962185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,8192,1.0628941853841145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,6144,0.8144044346279569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,6144,0.28394754727681476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,7168,1.028423097398546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,6144,0.8647768762376574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,65536,9.27536095513238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,7168,0.9129840003119575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,5120,0.6935795678032769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,4096,0.5079813533359104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,5120,0.7263786527845594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,6144,0.7595893541971842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,4096,0.19989421632554796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,5120,0.6358568933275011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,3584,0.43445422914293075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,4096,0.5830657747056749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,3584,0.18110666010114881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,4096,0.5172186427646214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,3584,0.5328853395250108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,3072,0.39161067538791233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,5120,0.2444373236762153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,3072,0.15897688600752088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,2560,0.3107279936472575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,3072,0.42235644658406574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,3584,0.45455111397637266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,2560,0.1420853270424737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,2048,0.2638631131913927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,2560,0.3797777758704291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,3072,0.4011528756883409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,2048,0.12168444527520074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,2560,0.336500432756212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,2048,0.2937591075897217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,1536,0.09793333212534587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,1536,0.24619023005167642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,2048,0.27820444107055664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,768,0.10505777597427368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,1024,0.07669422361585829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,1536,0.2184008889728122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,1024,0.1819902261098226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,1024,0.16411644882626003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,768,0.0682835578918457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,768,0.14351022243499756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,256,0.05837421947055393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,512,0.06096622016694811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,512,0.12222399976518418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,768,0.13662844234042698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,128,0.05095644460784065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,256,0.10840266942977905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,256,0.055345777008268565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,1024,0.15500799814860025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,128,0.052833778990639575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,64,0.044136888451046415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,128,0.10267555713653564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,256,0.08802577522065903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,32,0.04504266712400648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,128,0.06726844443215264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,64,0.1039208902253045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,512,0.08238666587405734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,32,0.10484977563222249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,512,0.11140799522399902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,65536,1.0065946578979492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,1536,0.21769066651662192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,51200,2.1492996215820312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,65536,2.760009765625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,16384,0.6457937558492025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,51200,0.8252124256557889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,16384,0.2377822134229872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,16384,0.8122231165568033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,12288,0.6380782127380371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,51200,3.4965928395589194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,65536,4.093382941351996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,16384,0.6718391312493218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,12288,0.5771359867519802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,12288,0.2066284418106079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,10240,0.46426041920979816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,10240,0.15665422545539007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,12288,0.49265156851874453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,8192,0.4111502170562744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,51200,2.3396292792426214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,10240,0.4880933231777615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,65536,3.0500365363227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,8192,0.1243528922398885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,10240,0.42081867323981387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,7168,0.3643724388546414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,8192,0.5335920121934679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,7168,0.12152177757687038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,6144,0.25233689943949383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,7168,0.41002665625678164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,6144,0.09877777761883205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,8192,0.3381964365641276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,5120,0.2385866641998291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,7168,0.3004968961079915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,6144,0.28043821122911244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,5120,0.08494755294587876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,5120,0.25847911834716797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,3584,0.16648356119791666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,4096,0.07095644209120008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,6144,0.373656882180108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,5120,0.21545510821872285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,3072,0.13656177785661486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,3584,0.06361866659588285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,3584,0.16938932736714682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,2560,0.09972977638244629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,3072,0.05659288830227322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,3584,0.1605315605799357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,3072,0.15602044264475504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,2560,0.05070755547947354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,2560,0.12415999836391872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,2048,0.09362488985061646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,4096,0.17846222718556723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,2560,0.11896533436245388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,2048,0.10195200310813056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,2048,0.04320444332228767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,4096,0.19447110758887398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,2048,0.10241955518722534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,1024,0.05083822210629781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,4096,0.17894132932027182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,1536,0.0354417794280582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,1536,0.08355022139019436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,3072,0.1438035567601522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,1024,0.06287733051511976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,1024,0.06504622432920668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,768,0.050273779365751475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,512,0.02914488977856106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,768,0.027456000447273254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,768,0.05481333202785916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,1536,0.0702915522787306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,256,0.019886222150590684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,512,0.04311466548177931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,512,0.025267556309700012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,1536,0.08134311437606812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,128,0.01832088828086853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,512,0.04559466573927137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,256,0.03831111060248481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,1024,0.029935999049080744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,64,0.015480889214409722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,128,0.036483556032180786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,32,0.015376889043384127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,128,0.02791022260983785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,64,0.03665688965055678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,32,0.03695111142264472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,256,0.022896000080638464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,768,0.038375112745496966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,128,0.02203911046187083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,65536,0.7495155334472656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,65536,2.149117363823785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,256,0.036347554789649114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,51200,1.6065048641628688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,16384,0.5411581993103027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,51200,0.5947617954678005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,16384,0.5931253433227539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,16384,0.18771466943952772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,65536,3.2487244076199002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,12288,0.4813635614183214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,16384,0.5125688976711696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,51200,2.408697764078776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,12288,0.14758755101097956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,12288,0.42932889196607804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,10240,0.3794204394022624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,65536,2.3501983218722873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,10240,0.11894844637976752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,12288,0.3865813414255778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,10240,0.36662311024136013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,8192,0.09722310966915554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,10240,0.32701065805223256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,7168,0.23386756579081217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,8192,0.31452978981865776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,8192,0.27782222959730357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,7168,0.2501715554131402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,6144,0.23069421450297037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,6144,0.07600533299975924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,7168,0.2273644341362847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,5120,0.17692800362904867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,6144,0.20165333482954237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,51200,1.8054478963216145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,5120,0.06531377633412679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,5120,0.1810213327407837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,8192,0.24398133489820692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,4096,0.14037599828508165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,3584,0.10692266623179118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,4096,0.05479022198253208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,4096,0.1428204509947035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,3584,0.12839199437035456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,3584,0.04893777767817179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,3072,0.09046044614579941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,7168,0.08713333474265204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,4096,0.13954222202301025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,2560,0.08993066681755914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,3072,0.04396355483267042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,3584,0.12448710865444607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,3072,0.1113448937733968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,2048,0.07095377975040011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,2560,0.09638310803307427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,2560,0.039107554488711886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,3072,0.1114657719930013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,1536,0.05219911204444038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,2048,0.0339955555068122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,2048,0.07886844211154513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,2560,0.09652888774871826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,1024,0.0354053344991472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,1536,0.02734755476315816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,2048,0.08124355475107829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,5120,0.1667004426320394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,768,0.029269334342744615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,1024,0.022820444570647344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,1024,0.04824089010556539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,1536,0.06565955612394545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,512,0.02001422146956126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,768,0.020664888951513503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,6144,0.18909422556559244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,768,0.03908533189031813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,1024,0.051226665576299034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,256,0.015576889117558798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,512,0.018933332628673978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,512,0.031721777386135526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,768,0.04407644602987501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,128,0.014535999960369535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,256,0.028280887338850234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,256,0.01752355529202355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,64,0.012245333029164208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,512,0.0373599992858039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,32,0.012426666915416718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,128,0.017349332571029663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,256,0.02970577942000495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,128,0.022602667411168415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,64,0.02717777755525377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,32,0.02699733277161916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,1536,0.06339466571807861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,65536,0.6673119862874349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,128,0.02716444432735443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,51200,1.4606346554226344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,65536,2.02164798312717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,51200,0.5070328712463379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,16384,0.5070115725199381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,51200,2.12776968214247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,65536,2.7284683651394315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,16384,0.15757866700490317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,16384,0.5326328807406956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,12288,0.3437528875139024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,16384,0.43932800822787815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,12288,0.1202382246653239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,10240,0.2687644428677029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,12288,0.3899128966861301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,8192,0.2160408894220988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,12288,0.32672354910108775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,10240,0.10081244177288479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,10240,0.3050319883558485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,7168,0.197580443488227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,51200,1.5155235926310222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,8192,0.24287467532687715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,10240,0.2731315559811062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,8192,0.08237510919570923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,6144,0.17278488477071127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,7168,0.08348177538977729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,7168,0.2121422158347236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,5120,0.15364178021748862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,6144,0.06505333052741157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,6144,0.18326844109429252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,7168,0.200217776828342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,4096,0.11903644932640924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,5120,0.05601155757904053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,8192,0.22055910693274605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,5120,0.15352977646721735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,65536,1.9817644755045574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,3584,0.10167821910646226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,4096,0.04726755619049072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,6144,0.16940977838304308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,4096,0.12591555383470324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,5120,0.14798488881852892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,3072,0.08945510784784953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,3584,0.10827110873328315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,2560,0.07024444474114312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,4096,0.11943644947475857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,3584,0.04176888863245646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,3072,0.03752266698413425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,2048,0.05810666746563382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,3584,0.1068355573548211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,2560,0.03342399994532267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,1536,0.04485155476464165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,3072,0.09535200066036648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,2048,0.028612444798151653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,2048,0.06749244530995686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,1024,0.02933599882655674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,2560,0.08200444115532769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,1536,0.02380888826317257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,1536,0.053972442944844566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,768,0.022421333524915908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,1024,0.020233778489960563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,1024,0.041160888142055936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,2048,0.07160622543758817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,1024,0.04528177777926127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,768,0.033815112378862165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,768,0.018180444836616516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,512,0.0176862229903539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,256,0.013471999930010902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,1536,0.05840177668465508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,512,0.016188444362746347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,512,0.02743822170628442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,768,0.038520889149771795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,128,0.010696888797812991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,256,0.015105777316623263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,512,0.03256888853179084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,2560,0.0820631119940016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,128,0.015241778559154935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,3072,0.09636622005038792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,256,0.024280000064108107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,32,0.010096000300513374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,256,0.025192888246642217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,128,0.02370044423474206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,128,0.019631110959582858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,64,0.023383110761642456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,32,0.023177777727444965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,65536,0.52347199122111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,64,0.009871111147933537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,51200,0.9865741729736328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,65536,1.3900498284233942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,16384,0.3615644507937961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,51200,0.4264390998416477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,51200,1.663052452935113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,16384,0.12729156017303467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,16384,0.42151112026638454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,65536,2.185434765285916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,12288,0.266973336537679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,12288,0.10501777463489109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,16384,0.359116448296441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,10240,0.21412000391218397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,12288,0.3169519901275635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,65536,1.6092888514200847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,8192,0.20246399773491752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,12288,0.27016178766886395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,10240,0.2605946593814426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,51200,1.24380980597602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,8192,0.06854844755596586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,7168,0.16670400566524932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,10240,0.2293919987148709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,8192,0.23436888058980307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,6144,0.15044532881842718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,7168,0.06130133072535197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,7168,0.17207466231452095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,8192,0.19152000215318468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,5120,0.10040533542633057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,6144,0.05333066648907132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,6144,0.14984799755944148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,4096,0.09290488560994466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,7168,0.17563733789655897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,5120,0.046048889557520546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,5120,0.12486222055223252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,6144,0.14305333296457926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,3584,0.0793164438671536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,4096,0.039373334911134504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,4096,0.10111377636591594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,5120,0.12606933381822374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,3072,0.0689306656519572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,3584,0.03450489044189453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,10240,0.08684800068537395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,3584,0.08926933341556126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,2560,0.05728889173931546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,4096,0.10217244757546319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,2048,0.04710311028692457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,3072,0.07847022347980075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,2560,0.02825333343611823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,3584,0.0905546678437127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,2560,0.0684444440735711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,1536,0.03727911247147454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,3072,0.07969955603281657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,2048,0.02427822185887231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,1024,0.02469066613250309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,2048,0.05673066775004069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,2560,0.07006755802366468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,1536,0.020844444632530212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,1536,0.04547911220126682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,768,0.02014666630162133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,2048,0.0595324436823527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,1024,0.017125333348910015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,512,0.014484445254007975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,1024,0.03511733478969998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,768,0.015810666812790763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,768,0.02864977717399597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,1024,0.039048890272776283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,256,0.0107351111041175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,512,0.01442399952146742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,512,0.02296177711751726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,768,0.033207999335394964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,128,0.00983911090426975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,256,0.01330311099688212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,512,0.028016000986099243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,3072,0.030964444080988567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,64,0.008442666795518663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,32,0.008617777791288164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,128,0.01351466692156262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,256,0.022135999467637803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,128,0.017606221967273288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,64,0.01961688862906562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,32,0.019482667247454327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,1536,0.048957334624396436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,65536,0.5023360252380371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,256,0.020590222544140287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,51200,0.9150773154364692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,65536,1.2920293807983398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,128,0.01956888867749108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,16384,0.3221520053015815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,51200,0.39302934540642637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,16384,0.3612364398108588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,16384,0.1149066686630249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,65536,1.8901244269476996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,16384,0.3197724554273817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,12288,0.26454488436381024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,51200,1.3172826766967773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,12288,0.08664621909459432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,10240,0.23613866170247397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,12288,0.25963110393948025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,65536,1.422760009765625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,10240,0.23932178815205893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,8192,0.15819999906751844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,12288,0.24446932474772134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,51200,1.1001858181423612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,8192,0.06144977940453423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,8192,0.21667467223273384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,10240,0.20336711406707764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,6144,0.13512355751461452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,7168,0.05749155415429009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,8192,0.16618933942582872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,7168,0.16984000470903185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,5120,0.12185066276126438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,7168,0.14653600586785212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,6144,0.04841422372394138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,6144,0.13509866926405165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,4096,0.09307555357615153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,5120,0.04166133205095927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,6144,0.12874755594465467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,5120,0.1132497787475586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,10240,0.08259555366304186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,4096,0.03462400039037069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,3584,0.07951377497778998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,5120,0.10947910944620769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,4096,0.09143466419643825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,3072,0.06920266813702054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,7168,0.14094221591949463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,3584,0.03053511182467143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,4096,0.0918053322368198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,3584,0.08033777607811822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,2560,0.054614222711986966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,3072,0.02852977646721734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,3072,0.07074577940834893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,2048,0.04686489038997226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,2560,0.025243555506070454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,3584,0.08243555492824979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,1536,0.035051554441452026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,2560,0.061443554030524365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,3072,0.07293422354592217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,2048,0.021744888689782884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,1024,0.02316266629430983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,2560,0.06315910816192627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,1536,0.018586667047606576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,1536,0.04043111205101013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,768,0.017441777719391715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,2048,0.05481600099139743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,1024,0.01573422220018175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,1024,0.03136266602410211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,1536,0.045307556788126625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,512,0.015219555960761176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,768,0.01460533340771993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,1024,0.03524000114864773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,256,0.011704000333944956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,768,0.026058667235904272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,512,0.013329777452680798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,512,0.02091466718249851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,128,0.009323555562231276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,256,0.0123857781291008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,256,0.018607111440764535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,512,0.025776000486479864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,64,0.00793422261873881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,128,0.01240533341964086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,32,0.008167111211352879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,256,0.02161688937081231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,128,0.017802667286660936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,64,0.017909333109855652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,128,0.017156443662113614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,32,0.01780177818404304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,65536,0.4155306551191542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,768,0.030597332451078627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,51200,0.7392969131469727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,65536,1.1910621854994032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,16384,0.26811377207438153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,2048,0.05087733268737793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,51200,0.3330942259894477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,65536,1.517784012688531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,16384,0.44278844197591144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,16384,0.09940977891286214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,65536,1.2516711552937825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,12288,0.07650666766696505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,12288,0.23708976639641655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,16384,0.30085333188374835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,10240,0.17907378408643934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,8192,0.1416133377287123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,12288,0.21546133359273276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,51200,0.9484124713473849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,10240,0.1896782186296251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,7168,0.13086756070454916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,8192,0.05981422132915921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,10240,0.17928266525268555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,8192,0.1904062165154351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,51200,1.1703689363267686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,6144,0.10430400239096747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,7168,0.04708710975117154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,8192,0.14643822775946722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,12288,0.2179795503616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,5120,0.08848266469107734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,6144,0.04098577631844415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,6144,0.11540888415442573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,7168,0.13643466101752386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,4096,0.07032089100943671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,5120,0.03576177689764235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,5120,0.0986257791519165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,6144,0.11507288614908855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,3584,0.0621982216835022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,10240,0.06636266575919257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,4096,0.07934755749172635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,3584,0.07133155398898654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,3072,0.05097777644793192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,3584,0.07311199771033393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,7168,0.14959200223286948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,3072,0.06283377938800387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,2560,0.04154133465554979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,3072,0.025598223010698955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,3072,0.06539911031723022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,2560,0.054392887486351855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,2048,0.03379199902216593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,4096,0.030805332793129817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,2560,0.0229093333085378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,5120,0.09786488612492879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,2560,0.05788889196183947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,1536,0.027560000618298847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,2048,0.020066666934225295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,3584,0.027519999278916255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,2048,0.0452133317788442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,1024,0.01962311069170634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,1536,0.036433776219685875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,1536,0.016957332690556843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,2048,0.04955644408861796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,768,0.015057777365048727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,1024,0.014532445205582513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,1024,0.027458666099442378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,1536,0.040708445840411715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,512,0.01145244472556644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,768,0.013481777575280933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,1024,0.03238755464553833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,256,0.008656000097592672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,512,0.012332444389661154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,512,0.018940443793932598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,768,0.02769244379467434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,256,0.011515555282433828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,256,0.01679111189312405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,512,0.023016000787417095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,64,0.006647111227114995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,256,0.020679111282030743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,128,0.01593866613176134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,128,0.011555555793974133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,4096,0.08261777957280476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,64,0.015950222810109455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,128,0.016511110795868766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,32,0.015847999188635085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,768,0.023457777169015672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,65536,0.3837591012318929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,128,0.007600000335110559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,65536,0.9618470933702258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,51200,0.6754150920444064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,16384,0.2603253258599175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,32,0.006794666664467917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,51200,0.2982986768086751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,16384,0.08557599782943726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,16384,0.2645466592576769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,12288,0.19621511300404867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,65536,1.0395581987169054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,51200,0.9548941718207465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,12288,0.07153955433103773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,12288,0.1932506693734063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,10240,0.1758319934209188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,16384,0.2664462195502387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,8192,0.14191110928853354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,10240,0.05739466349283854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,12288,0.18635643853081596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,10240,0.1671217812432183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,7168,0.110225776831309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,8192,0.1309946642981635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,8192,0.04567822151713901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,10240,0.1612702210744222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,6144,0.10927910937203301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,7168,0.04108000132772658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,7168,0.11658577124277751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,8192,0.1660604476928711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,65536,1.2204408645629883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,5120,0.08133955796559651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,6144,0.036646223730511136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,6144,0.10083199871910943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,4096,0.0702257752418518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,7168,0.1145484447479248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,5120,0.03251466817326016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,6144,0.10184533066219753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,3584,0.05948533614476522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,5120,0.08722755644056533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,4096,0.026877333720525105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,4096,0.06899022393756442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,3072,0.04860711097717285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,5120,0.0871351096365187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,3584,0.02532711128393809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,51200,0.7758328649732801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,3584,0.06164533562130398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,4096,0.0716222193506029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,3072,0.02234755622016059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,3072,0.054097778267330594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,3584,0.0659137765566508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,2048,0.0328382220533159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,2560,0.020420443680551317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,2560,0.04674222071965536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,3072,0.05779288874732124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,1536,0.025614221890767414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,2048,0.01755911111831665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,2048,0.03955199983384874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,2560,0.050534221861097545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,1024,0.019533332851197984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,1536,0.015487111277050443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,1024,0.02422844370206197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,1024,0.013073777986897362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,768,0.015704888436529372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,1024,0.028510222832361858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,768,0.020491555333137512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,2560,0.042749332057105176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,768,0.0122097780307134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,512,0.011083555718262991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,768,0.024417777856190998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,512,0.011372444530328115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,256,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,512,0.016536000702116225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,512,0.020999999509917364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,256,0.014787554740905762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,128,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,256,0.018505778577592637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,1536,0.031785776217778526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,64,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,128,0.014156444205178155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,128,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,1536,0.03619822197490268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,32,0.006018666757477655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,128,0.014463999205165438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,64,0.014158222410413953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,32,0.014157333307796054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,65536,0.7210311359829373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,65536,0.31293243832058376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,51200,0.5438035329182943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,256,0.01092711091041565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,2048,0.04334577918052673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,16384,0.17655732896592882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,65536,0.960685306125217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,51200,0.24278312259250215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,12288,0.13483999835120306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,16384,0.21469777160220674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,65536,0.8791191313001844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,12288,0.05684533384111193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,12288,0.1615404420428806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,16384,0.20790489514668783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,10240,0.11912088923984104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,51200,0.8831786579555936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,8192,0.0964168906211853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,51200,0.6713448630438911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,10240,0.04829511046409607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,12288,0.16844711038801405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,10240,0.13583732975853816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,7168,0.08711466524336074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,8192,0.04316355453597175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,8192,0.1223555538389418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,6144,0.07316088676452637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,10240,0.13605066140492758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,7168,0.03492177857293023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,7168,0.09696355793211196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,5120,0.05896355708440145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,8192,0.11919289165072972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,6144,0.03138755427466498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,7168,0.09877955251269871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,4096,0.046781334612104625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,6144,0.08843200074301825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,16384,0.07851733101738824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,5120,0.02827199962404039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,5120,0.07079111205206977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,3584,0.040234665075937905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,6144,0.08973689211739434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,4096,0.023574221465322707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,3072,0.036179555786980525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,4096,0.0580328901608785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,5120,0.07442577679951985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,3584,0.02164177762137519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,2560,0.02918222215440538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,3584,0.052925331725014575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,4096,0.06367555591795179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,3072,0.019879110985332064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,2048,0.024021334118313257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,3072,0.04605244596799215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,2560,0.01829866733815935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,1536,0.0195448895295461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,2560,0.039896888865364924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,2048,0.015782222151756287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,3072,0.05009777678383721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,2048,0.03398666779200236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,1024,0.014226666755146451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,1536,0.013815110756291283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,2048,0.037998222642474704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,1536,0.027082666754722595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,768,0.011433777709801992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,1024,0.011951111257076263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,1024,0.02072266737620036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,1536,0.03180711136923896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,512,0.008815999660227034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,768,0.017650667164060805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,1024,0.025176889366573755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,512,0.014224888549910652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,256,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,512,0.010567111273606619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,512,0.01893422173129188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,3584,0.05693866809209188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,256,0.012776888906955719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,256,0.009752000371615091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,256,0.016456888781653512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,2560,0.044250667095184326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,64,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,128,0.012422222230169507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,32,0.0053031109273433685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,128,0.010149333212110732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,64,0.0123857781291008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,128,0.013460444079505073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,768,0.011034666664070554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,768,0.02165066699186961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,65536,0.572455988989936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,65536,0.2896631028917101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,128,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,51200,0.47742308510674375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,16384,0.1759199963675605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,51200,0.22907466358608672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,65536,0.8994248708089193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,12288,0.13047022289699978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,51200,0.6915715535481771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,16384,0.07191911008622912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,16384,0.2167431116104126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,65536,0.8159448835584852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,32,0.012423999607563019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,10240,0.12015555964575873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,12288,0.055083556307686694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,51200,0.6411689122517904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,16384,0.20663377973768446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,8192,0.09618577692243789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,10240,0.04790755444102817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,12288,0.14949156178368464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,10240,0.13064177831013998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,7168,0.08534310923682319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,8192,0.0377493335141076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,6144,0.0726426641146342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,7168,0.03514400124549866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,8192,0.1031564474105835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,10240,0.1328995492723253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,7168,0.08854222297668457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,5120,0.057393776045905225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,8192,0.10964266459147136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,6144,0.03046400017208523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,7168,0.09928622510698105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,6144,0.07640888955858019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,5120,0.02649333410792881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,5120,0.0648711125055949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,6144,0.08602311213811238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,3584,0.04007822275161743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,4096,0.02278577784697215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,5120,0.07314933008617826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,4096,0.05292977889378866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,3072,0.034817778401904635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,3584,0.020641777250501845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,4096,0.061623109711541064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,3584,0.04751911097102695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,2560,0.029147555430730183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,3072,0.01867377758026123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,3072,0.04202311237653097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,3584,0.05567110909356011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,2048,0.02386133372783661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,12288,0.15958666801452637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,2560,0.017116443978415597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,2560,0.03676533367898729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,1536,0.019222221440739103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,2048,0.015136889285511441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,2048,0.0307004451751709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,1024,0.014656000667148165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,1536,0.013520888984203339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,1536,0.024142222272025213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,2048,0.0376782218615214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,4096,0.045401778486039907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,768,0.01179377817445331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,1536,0.030449777841567993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,1024,0.018995554910765756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,512,0.008634666601816813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,768,0.010979555547237396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,768,0.016368889146380954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,1024,0.024197333388858374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,256,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,512,0.013369777964221107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,768,0.021448888712459143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,256,0.009746666583749983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,256,0.01204622205760744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,128,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,512,0.0185208883550432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,3072,0.049192888869179614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,64,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,256,0.016494222813182406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,128,0.01000088867213991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,128,0.011420444481902652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,2560,0.04353777898682488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,64,0.01146755533085929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,128,0.012409778104888068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,1024,0.011767110890812345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,32,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,65536,0.5341013272603353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,512,0.010335110955768162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,65536,0.2541617817348904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,51200,0.42547199461195206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,16384,0.1417235533396403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,65536,0.7523262235853406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,51200,0.19940355088975695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,32,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,65536,0.6857057677374946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,16384,0.07039199935065375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,16384,0.18702933523390028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,12288,0.10725333293279012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,51200,0.6501368946499294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,10240,0.09672888782289292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,12288,0.05351555678579542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,16384,0.17446577548980713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,51200,0.5344124370151097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,12288,0.1461084418826633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,10240,0.04552177919281853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,8192,0.07688533597522311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,10240,0.11417955822414821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,12288,0.13460177845425075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,7168,0.06607288784450956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,8192,0.03564888901180691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,10240,0.11321511533525254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,6144,0.058456891112857394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,8192,0.09001066949632432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,7168,0.03282399972279867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,7168,0.07810399929682414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,5120,0.04728177852100796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,8192,0.09517777628368801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,6144,0.029013332393434312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,4096,0.035951998498704694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,7168,0.0840488870938619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,5120,0.025614221890767414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,5120,0.05755111243989733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,6144,0.07171466615464953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,3584,0.031192888816197712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,4096,0.021754667162895203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,4096,0.04728177852100796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,3072,0.028365333875020344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,3584,0.019687111179033916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,4096,0.053489777776930064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,2560,0.02238666680124071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,3072,0.017638223038779367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,3072,0.03774666786193848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,3072,0.042747557163238525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,2560,0.03283022178543939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,2048,0.018791110979186166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,2560,0.0376853346824646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,2048,0.027344889110989038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,6144,0.06710666418075562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,1536,0.015042665931913586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,2048,0.014567111929257711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,2048,0.03271911210483975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,1024,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,5120,0.06252533197402954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,1536,0.012153777811262341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,1536,0.022023111581802368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,768,0.008921778036488427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,1536,0.026758222116364375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,1024,0.010692444112565784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,1024,0.017461333009931777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,512,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,768,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,3584,0.04787733488612705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,1024,0.02179199953873952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,256,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,512,0.012146666646003723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,512,0.009555555880069733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,768,0.018998222218619455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,128,0.004960888789759742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,256,0.009345778160625035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,256,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,512,0.01792355544037289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,64,0.00453333349691497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,128,0.010623111493057676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,128,0.009071110851234859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,32,0.004604444321658877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,64,0.010631110933091907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,128,0.012376888758606382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,32,0.010679110884666443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,3584,0.04240444302558899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,65536,0.44316355387369794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,768,0.014946666028764514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,65536,0.23131023512946236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,2560,0.01627200014061398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,51200,0.383424891365899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,16384,0.11175644397735596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,65536,0.6611857944064671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,51200,0.18863022327423096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,12288,0.0944746666484409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,51200,0.5239004558987087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,16384,0.05746044715245565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,65536,0.6545137829250759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,12288,0.04462755719820658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,51200,0.5159689055548774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,12288,0.11461244689093696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,16384,0.1705555518468221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,10240,0.08722577492396037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,8192,0.06803466876347859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,12288,0.1315635575188531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,10240,0.038718223571777344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,10240,0.09457333220375909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,7168,0.05750399827957153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,8192,0.0306915541489919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,8192,0.07864266633987427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,7168,0.02808444367514716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,8192,0.09132977988984849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,7168,0.06794489092297025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,6144,0.04878044459554884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,16384,0.16460711426205105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,256,0.01519555515713162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,5120,0.04051733348104689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,7168,0.08125422398249309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,6144,0.025443555580245122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,4096,0.03330488999684652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,6144,0.059474666913350425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,5120,0.021925333473417494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,5120,0.05068000157674154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,6144,0.07155022356245253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,3584,0.029488888051774766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,4096,0.018698667486508686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,4096,0.04219733344184028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,5120,0.06098488966623942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,3072,0.025814221964942083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,3584,0.017322666115230985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,3584,0.03765866822666592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,4096,0.05125155713823107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,3072,0.01606399979856279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,3072,0.033415112230512835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,3584,0.04667733444107903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,2560,0.014552889598740472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,2560,0.028883556524912517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,3072,0.041456000672446355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,1536,0.014931556251313953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,2048,0.013354666531085968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,2048,0.024513777759340074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,2560,0.036213331752353244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,1024,0.011234666738245221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,1536,0.019733332925372653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,1536,0.011773332953453064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,2048,0.031698667340808444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,768,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,1024,0.010324444207880232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,1536,0.026015111141734656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,1024,0.020968889196713764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,768,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,768,0.013518222504191928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,512,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,2560,0.022652443912294176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,256,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,10240,0.11149244838290745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,512,0.009377777576446533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,2048,0.018608889646000333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,128,0.004652444273233414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,512,0.011018666956159802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,256,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,256,0.010088000032636855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,64,0.004218666503826777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,512,0.016178665889634024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,32,0.004341333276695675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,128,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,128,0.009729777773221334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,256,0.014278221461508008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,64,0.009726222190592024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,32,0.009654222263230218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,1024,0.01569600072171953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,65536,0.1762453317642212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,51200,0.2538906733194987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,65536,0.5212231212192112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,768,0.0188755558596717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,16384,0.10132444567150539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,51200,0.140374223391215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,51200,0.40713776482476133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,65536,0.5180631213717991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,12288,0.07493066787719727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,16384,0.04533066683345371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,16384,0.12778222560882568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,10240,0.06149599949518839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,12288,0.03474666674931844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,51200,0.40725866953531903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,12288,0.10363199975755479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,16384,0.1397155523300171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,8192,0.04804799954096476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,10240,0.03319022059440613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,65536,0.34381688965691465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,12288,0.10631022188398574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,7168,0.043529778718948364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,10240,0.08075200186835395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,8192,0.06670133272806804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,8192,0.025153777665562097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,6144,0.03495288888613383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,7168,0.02223644488387638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,10240,0.08951822254392837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,7168,0.058550225363837353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,5120,0.029940442906485662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,6144,0.05179644293255276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,6144,0.020106666617923312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,5120,0.017840888765123155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,5120,0.04316622349951002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,6144,0.05890933010313246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,4096,0.02536800007025401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,5120,0.05115288827154371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,128,0.012091555529170565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,3584,0.021380444367726643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,4096,0.0357582237985399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,4096,0.042903112040625684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,3584,0.032347556617524885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,3072,0.018933332628673978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,3584,0.03875911235809326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,8192,0.07457066906823052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,2560,0.016287111573749118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,3072,0.028664890262815688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,3072,0.013401777380042605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,7168,0.06747111346986559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,3072,0.034401777717802264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,2048,0.013853333062595792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,2560,0.012326222327020435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,2560,0.02460177739461263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,1536,0.011048888994587792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,2048,0.011057777537239922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,2560,0.03087288803524441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,2048,0.021306667062971327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,1024,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,4096,0.015577778220176697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,1536,0.010057777994208867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,1536,0.017243555850452848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,768,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,1024,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,1024,0.013576000101036496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,3584,0.014544889330863953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,512,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,1536,0.022064889470736187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,768,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,768,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,256,0.004544888933499654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,1024,0.018181333939234417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,512,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,512,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,128,0.003908444609906939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,768,0.016176000237464905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,256,0.008057778080304464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,256,0.008991999758614434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,64,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,512,0.01460000044769711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,32,0.00361333332127995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,128,0.008055110772450766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,128,0.008712000317043727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,256,0.013101333545313941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,64,0.008662222160233391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,32,0.008734222087595198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,65536,0.2889297803243001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,2048,0.0262044436401791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,65536,0.162207113371955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,51200,0.21910756164126924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,16384,0.07663644684685601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,65536,0.41828711827596027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,51200,0.1277351114484999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,51200,0.32994577619764537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,12288,0.06013244390487671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,16384,0.04361244373851352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,16384,0.10668088992436726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,10240,0.049124442868762545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,65536,0.5046462482876247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,12288,0.03414933217896356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,51200,0.4079271157582601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,12288,0.08016977707544963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,16384,0.13550933202107748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,10240,0.030596445004145306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,10240,0.06863022512859769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,12288,0.10433689090940688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,7168,0.033129778173234724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,8192,0.023631999890009563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,128,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,8192,0.0549759997261895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,10240,0.08817155493630303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,6144,0.026341333985328674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,7168,0.022102221846580505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,7168,0.04904266529613071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,8192,0.07266666491826375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,6144,0.019493333167499967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,6144,0.04268977708286709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,7168,0.06532533301247491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,4096,0.01976355579164293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,5120,0.017083555459976196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,5120,0.03644799855020311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,6144,0.05744000275929769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,3584,0.017113778326246474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,4096,0.015125332607163323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,4096,0.03029511041111416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,5120,0.049364444282319814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,3072,0.014885332849290637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,3584,0.014209777116775513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,3584,0.027624888552559748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,8192,0.03905511233541701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,2560,0.013209777573744455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,3072,0.013330666555298699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,3584,0.037610666619406805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,3072,0.024263999528355066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,2048,0.011048888994587792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,2560,0.012176888684431711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,2560,0.02123466630776723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,5120,0.023344889283180237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,2048,0.011060444845093621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,2048,0.018326222896575928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,2560,0.029190222422281902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,1536,0.010013333625263637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,2048,0.025236444340811834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,1024,0.007093333535724216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,1536,0.015171556009186639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,1024,0.009341333475377824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,1536,0.021544888615608215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,1024,0.011961778004964193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,768,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,4096,0.041731556256612144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,512,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,768,0.00903733323017756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,768,0.010450666977299584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,768,0.016189333465364244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,512,0.009039999710188972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,512,0.008291555775536431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,1536,0.00885688927438524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,256,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,256,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,256,0.012994666894276937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,128,0.00793511089351442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,64,0.0031848889258172777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,128,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,32,0.0032835555159383347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,1024,0.0181351105372111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,128,0.01017955525053872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,64,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,65536,0.1778479946984185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,32,0.007948444121413762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,3072,0.03404355711407132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,51200,0.16292177306281194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,512,0.014468444718254937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,65536,0.11652444468604194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,16384,0.05328266819318136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,128,0.003536000019974179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,51200,0.09295644362767537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,51200,0.25241155094570583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,12288,0.04107022285461426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,65536,0.3669271204206679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,16384,0.03208444515864054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,16384,0.08403644296858047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,10240,0.033153778976864286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,51200,0.2867617872026232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,12288,0.025016888976097107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,16384,0.10167911317613389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,8192,0.025664889150195654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,10240,0.022509333160188463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,10240,0.053470220830705434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,12288,0.07698844538794623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,7168,0.022431999444961548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,256,0.0038888889054457345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,8192,0.01739199956258138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,8192,0.04401422209209866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,10240,0.06526755624347262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,6144,0.01889155473974016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,7168,0.016193777322769165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,7168,0.039056890540652804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,8192,0.055290665891435414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,5120,0.01612355477280087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,6144,0.014785778191354541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,6144,0.034338666333092585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,7168,0.049286223120159574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,4096,0.013447110851605734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,5120,0.013256000147925483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,6144,0.043750223186280995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,5120,0.029678222205903795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,3584,0.012158222496509552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,4096,0.011755555868148804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,4096,0.0247688889503479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,5120,0.037963555918799505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,3072,0.010942222343550788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,12288,0.063137776321835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,3584,0.022072000635994807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,3584,0.011323555476135679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,2560,0.009378666679064432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,3072,0.010701333483060202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,3584,0.028712001111772325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,2048,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,2560,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,2560,0.01757333344883389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,3072,0.026072000463803608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,1536,0.006737777756320105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,2560,0.023391111029518977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,2048,0.015303111738628812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,65536,0.319942209455702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,1024,0.005434666656785541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,1536,0.012617777619096967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,1536,0.008886222210195329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,2048,0.020454222957293194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,1024,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,1536,0.01761155492729611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,768,0.004575110971927643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,1024,0.014825777875052558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,4096,0.03166133496496413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,512,0.0038497779104444715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,768,0.009103111094898647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,768,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,3072,0.019955555597941082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,256,0.00318755561278926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,768,0.014177777700954013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,512,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,512,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,128,0.002833777831660377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,2048,0.009504888620641496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,256,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,512,0.012534221841229333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,64,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,128,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,128,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,256,0.011383111278216044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,32,0.002826666666401757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,64,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,65536,0.15938133663601345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,1024,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,128,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,32,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,51200,0.13216977649264866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,65536,0.2784106731414795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,16384,0.045886221859190196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,51200,0.09056799941592747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,51200,0.2209084431330363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,65536,0.3612622155083551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,16384,0.031123555368847314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,16384,0.07261421945359972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,12288,0.03279555506176419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,51200,0.28498755560980904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,256,0.006989333364698622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,10240,0.02539644473128849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,12288,0.024522667129834492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,12288,0.05533155467775133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,8192,0.020448888341585796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,10240,0.022300443715519373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,10240,0.04657422171698677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,7168,0.017451556192504037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,12288,0.07541155815124512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,8192,0.01732711162832048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,10240,0.06556444697909884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,8192,0.03857333461443583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,6144,0.015375110838148328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,65536,0.11415821976131862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,7168,0.01609599921438429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,7168,0.03439022103945414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,5120,0.013398222625255585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,8192,0.05436533358361986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,7168,0.047566221819983594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,6144,0.014552889598740472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,6144,0.030890666776233252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,4096,0.01164177805185318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,5120,0.013292444248994192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,3584,0.010263111028406356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,5120,0.025824889540672302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,6144,0.04308266772164238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,4096,0.011759999725553723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,3072,0.00903555585278405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,5120,0.03719377848837111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,3584,0.019713777634832595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,16384,0.10010400083329941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,2560,0.007748444047239091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,3072,0.017906667457686532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,3072,0.010724444356229572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,3584,0.027894222074084814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,2048,0.006690666493442323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,2560,0.010044444766309526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,2560,0.015489778584904142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,2048,0.013575110998418597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,2048,0.00945955514907837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,1536,0.005800888770156437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,2048,0.020216888851589628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,1536,0.011393778026103973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,1024,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,1536,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,1536,0.01718133356836107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,3584,0.011342222491900126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,768,0.004132444245947732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,1024,0.008295111358165741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,4096,0.03153955605294969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,768,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,768,0.008635555704434713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,1024,0.014527999692493014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,3072,0.02544622288809882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,768,0.013400889105266996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,256,0.002893333426780171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,512,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,512,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,2560,0.022661333282788593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,128,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,512,0.012104888757069906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,256,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,256,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,128,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,64,0.0027502222607533136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,256,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,128,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,32,0.002834666727317704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,4096,0.02167644434505039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,128,0.008770667016506195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,1024,0.009332444104883406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,65536,0.13307733005947536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,64,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,32,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,51200,0.1030684444639418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,512,0.0034977777136696708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,65536,0.11087999741236369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,16384,0.031968000862333514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,51200,0.08877955542670356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,65536,0.23361423280504015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,51200,0.18465867307451037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,12288,0.02330222229162852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,16384,0.03009066647953457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,16384,0.06145155429840088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,65536,0.34787466790941024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,10240,0.01878666712178124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,51200,0.2746346791585286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,12288,0.023578666978412207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,12288,0.04708800050947401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,16384,0.09662755330403645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,8192,0.014506666196717156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,10240,0.021155556042989094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,10240,0.03999377621544732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,7168,0.013073777986897362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,12288,0.07380355728997125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,8192,0.016565332810084026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,6144,0.011389333340856763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,8192,0.0332640012105306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,10240,0.06286133660210504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,7168,0.01518133282661438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,5120,0.010219555762079027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,7168,0.02978133492999607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,8192,0.05231377813551161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,6144,0.014001778430408902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,6144,0.026368889543745253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,4096,0.008604444563388824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,7168,0.0468408895863427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,5120,0.01274311128589842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,3584,0.00810222244924969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,6144,0.04023911224471198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,4096,0.011423110961914062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,4096,0.018929777873886954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,5120,0.03547999925083584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,3584,0.010790222220950656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,3584,0.01737866633468204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,4096,0.029934220843844946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,2560,0.006292444550328785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,3072,0.010354667074150508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,3072,0.01573511130279965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,2560,0.013912889692518445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,2560,0.009690666364298927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,2048,0.005632888939645555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,3072,0.024735111329290602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,2048,0.00924177798959944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,2560,0.022063111265500385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,1536,0.004591111093759537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,2048,0.012141333685980903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,1536,0.008546666966544257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,2048,0.01959288948112064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,1024,0.0038808890514903595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,5120,0.022755554980701868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,1536,0.010423111418883005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,768,0.0035217776894569397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,3072,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,1024,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,1536,0.016884444488419425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,512,0.003137777869900068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,1024,0.008656000097592672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,768,0.007903999752468532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,256,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,512,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,1024,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,3584,0.027122666438420612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,512,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,256,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,256,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,512,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,64,0.0023555554863479403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,128,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,128,0.0069955554273393415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,32,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,256,0.011342222491900126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,65536,0.07657511366738214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,128,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,64,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,32,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,51200,0.05960622098710802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,65536,0.19024533695644805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,65536,0.10977422528796726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,16384,0.02244088881545597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,768,0.007672888537247975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,51200,0.14926044146219888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,65536,0.3416897720760769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,128,0.002475555572244856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,16384,0.050549334949917264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,12288,0.01609244445959727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,16384,0.02939644455909729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,51200,0.2694639894697401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,10240,0.015601777368121676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,12288,0.03881866733233134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,12288,0.0717395544052124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,10240,0.020966221888860066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,8192,0.012774222426944308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,10240,0.06194577614466349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,51200,0.08859199947781032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,8192,0.027458666099442378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,7168,0.01130311108297772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,768,0.013802666631009845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,8192,0.01648177703221639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,8192,0.05114577876196968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,6144,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,7168,0.0151591118839052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,7168,0.024811555941899616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,12288,0.022687999738587275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,5120,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,6144,0.013768889009952545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,7168,0.04514044523239136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,10240,0.03326133224699233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,4096,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,6144,0.039212445418039955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,5120,0.01923644377125634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,4096,0.011336888704035016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,4096,0.01646133263905843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,3584,0.006648889018429651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,4096,0.02900622288386027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,3584,0.010703110860453712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,3072,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,3584,0.01514311134815216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,3584,0.02678044471475813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,3072,0.010323555933104621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,2560,0.005454222361246745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,3072,0.01368533323208491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,16384,0.09477511379453872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,2048,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,6144,0.02200800014866723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,3072,0.02403822210099962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,2560,0.009667555491129557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,1536,0.0041982221106688184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,2560,0.012159110771285163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,2048,0.00904444439543618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,5120,0.01255733354224099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,2048,0.0109120003051228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,1024,0.003544888976547453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,2560,0.02159822152720557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,1536,0.009289777941173976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,768,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,2048,0.01886133352915446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,5120,0.0341306659910414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,1024,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,512,0.0028320000403457214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,1024,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,1536,0.0166302224000295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,768,0.007615110940403408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,256,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,768,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,1024,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,512,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,512,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,128,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,256,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,256,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,512,0.01108266661564509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,64,0.0021342221233579847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,32,0.002191999927163124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,128,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,128,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,256,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,65536,0.0554017788834042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,64,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,128,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,32,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,51200,0.044010665681627065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,65536,0.110024889310201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,16384,0.012082666986518435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,65536,0.18879289097256133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,1536,0.008700444466537898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,51200,0.08842666943868001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,51200,0.1483831140730116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,12288,0.012860444684823355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,16384,0.028743111424975928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,16384,0.05017955435646904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,65536,0.3392808967166477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,10240,0.011687111523416309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,12288,0.022521777285469904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,51200,0.26540178722805446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,8192,0.012051555845472546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,12288,0.03841777642567953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,768,0.01276533305644989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,10240,0.02088800072669983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,7168,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,10240,0.03252089023590088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,12288,0.07160088751051161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,8192,0.016177778442700703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,8192,0.027179555760489568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,10240,0.06112533145480686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,6144,0.009436444275909001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,7168,0.015114666687117683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,7168,0.024115555816226538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,5120,0.008349333372380998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,8192,0.05071377754211426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,6144,0.013777777552604675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,6144,0.021661332911915247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,7168,0.04500000013245476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,4096,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,5120,0.012503110700183444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,5120,0.018927999668651156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,6144,0.03894311189651489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,3584,0.0063546668324205614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,4096,0.011400889191362592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,4096,0.016188444362746347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,5120,0.03404622276624044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,3072,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,3584,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,3584,0.014829332629839579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,4096,0.028759111960728962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,2560,0.005163555757866966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,3072,0.010246222217877706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,3072,0.013629333012633853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,2048,0.004515555583768421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,2560,0.012106666962305704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,3072,0.023698666029506262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,2560,0.020961778031455144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,2048,0.009031111167536842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,1536,0.0038942222793896994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,2048,0.010718222293588849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,1024,0.0032426667296224167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,1536,0.008470222353935242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,2048,0.018577777677112155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,1536,0.009089777866999308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,768,0.0028888889484935333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,1024,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,1024,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,1536,0.016499555773205228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,512,0.0025315555847353404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,768,0.007701333198282454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,1024,0.013737777868906656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,768,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,256,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,512,0.006041777630647023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,512,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,3584,0.02622577713595496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,128,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,768,0.012379555238617791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,256,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,256,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,64,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,512,0.011398221883508893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,32,0.002193777718477779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,128,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,128,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,65536,0.053215109639697604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,256,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,51200,0.04202311237653097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,64,0.005103999955786599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,16384,0.01516711049609714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,32,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,65536,0.18903199831644693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,51200,0.14783199628194174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,16384,0.05003289050526089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,8192,0.009882666998439366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,16384,0.09413777457343207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,7168,0.009105777574910058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,8192,0.02677333354949951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,6144,0.008436444732877944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,2560,0.009707555174827576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,5120,0.007761778102980719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,7168,0.02403733299838172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,4096,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,6144,0.02145511077509986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,3584,0.006862222320503659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,5120,0.018906666172875296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,12288,0.012473777764373355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,3072,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,4096,0.016187555260128446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,2560,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,2048,0.004322666674852371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,3584,0.014780445231331719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,1536,0.0038533334930737815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,10240,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,3072,0.013471999930010902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,1024,0.0032142222755485107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,2560,0.011988444460762871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,768,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,2048,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,512,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,256,0.002190222135848469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,1536,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,128,0.0018728888697094387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,10240,0.032441778315438166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,1024,0.007681777907742395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,64,0.0018959999498393799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,768,0.007176000210973952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,512,0.006082666830884085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,256,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,65536,0.04778933193948534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,128,0.005197333378924264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,64,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,51200,0.039424889617496066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,32,0.004969777746333016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,128,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,65536,0.1890204482608371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,12288,0.00849777791235182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,51200,0.14772355556488037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,16384,0.050069332122802734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,10240,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,8192,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,7168,0.006688000013430913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,12288,0.038141333394580416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,12288,0.03794755538304647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,6144,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,10240,0.03224710954560174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,8192,0.026837334036827087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,7168,0.02409066590997908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,4096,0.01074488874938753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,3584,0.010060444474220276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,6144,0.021538666552967493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,5120,0.018935999936527677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,3072,0.009168888959619734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,2560,0.006411555740568373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,4096,0.016177778442700703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,2048,0.004436444491147995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,1536,0.003842666745185852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,16384,0.010017777482668558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,1024,0.0032337777730491427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,3072,0.013541333377361298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,2560,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,512,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,2048,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,256,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,1536,0.00905511114332411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,128,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,1024,0.007731555236710443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,768,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,64,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,512,0.006009777800904379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,32,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,256,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,128,0.005167110926575131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,64,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,5120,0.006069333189063602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,32,0.004975111120276981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,3584,0.014833778142929077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,768,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,16384,0.8558150927225748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,12288,2.0288382636176214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,12288,0.7513795428805881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,16384,2.6842257181803384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,12288,2.166910171508789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,16384,3.0051146613227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,10240,0.5649946530659994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,10240,1.6509884728325739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,51200,3.266950183444553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,8192,0.4486106766594781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,10240,1.8934186299641926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,8192,1.3344186147054036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,7168,0.39879732661777073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,7168,1.2084666358100045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,8192,1.3989253573947484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,6144,0.399674654006958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,6144,1.1925599839952257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,7168,1.2284684711032443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,6144,0.9925119611952039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,5120,0.29810844527350533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,5120,0.8950177298651801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,4096,0.250636445151435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,5120,0.8231511116027832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,4096,0.6676506466335721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,3584,0.22353598806593153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,3584,0.665412425994873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,3072,0.4881413247850206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,3072,0.19453956021202934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,3584,0.5873093075222439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,3072,0.49958133697509766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,2560,0.5307795736524794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,2560,0.17214043935139975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,2048,0.32741954591539174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,2560,0.4912506739298503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,4096,0.8051333427429199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,1536,0.28367288907368976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,2048,0.3434720039367676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,1024,0.1805644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,51200,8.679695129394531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,1536,0.1150568856133355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,768,0.14207911491394043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,1024,0.09205333391825359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,1536,0.28099555439419216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,1024,0.2235786649915907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,512,0.10489688979254828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,256,0.07304088936911689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,512,0.07467999723222521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,512,0.1416435506608751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,128,0.05889422363705105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,256,0.06771822108162774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,256,0.12751289208730063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,768,0.16621866491105822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,64,0.05096977618005541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,32,0.05300800005594889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,2048,0.17478133572472465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,128,0.06457333432303534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,128,0.11940977308485244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,64,0.12127466996510823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,32,0.12127110693189834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,768,0.08086311154895358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,51200,11.972632514105904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,51200,2.6211920844184027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,16384,2.0885963439941406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,65536,3.318936030069987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,16384,2.2247581481933594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,16384,0.663315561082628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,12288,0.5293822288513184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,12288,1.562839084201389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,51200,6.382997300889756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,10240,1.4496613608466253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,12288,1.793060514662001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,10240,0.42921866310967344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,8192,0.3546284569634332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,10240,1.3966088824801977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,8192,0.9969057506985135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,7168,0.9531271192762586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,65536,8.29825676812066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,7168,0.32062578201293945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,8192,1.1002933714124892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,7168,0.988242679172092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,6144,0.7525573306613498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,5120,0.6118097835116917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,6144,0.27989599439832896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,5120,0.23732622464497885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,6144,0.8330488734775119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,4096,0.5127102004157172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,5120,0.6408551004197862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,4096,0.1980959971745809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,3584,0.42908265855577254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,51200,9.509880065917969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,4096,0.5309404267205132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,3584,0.17127822505103218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,3072,0.1530053350660536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,3072,0.38949865765041775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,3584,0.44921244515313047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,2560,0.1343031062020196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,2048,0.2611431015862359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,1536,0.19841421975029838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,2560,0.3317200077904595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,2048,0.26878311898973256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,1024,0.12675110499064127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,1536,0.09059022532569037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,1536,0.23164267010158965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,1024,0.07262133227454291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,768,0.11939377254909939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,512,0.08733333481682672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,1024,0.1616791089375814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,3072,0.3787244425879584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,768,0.06463111109203763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,768,0.13121244642469618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,256,0.05486399928728739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,512,0.05945689148373074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,128,0.04660711023542616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,256,0.05435644586881002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,512,0.11160443888770209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,2560,0.3246648841434055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,64,0.04080977704789903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,128,0.05171822177039253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,32,0.04114133450720046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,128,0.09458400143517388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,256,0.09986399941974217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,2048,0.11424267292022705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,64,0.09695644511116876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,32,0.09638222058614095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,65536,12.369249131944445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,16384,0.7122737566630045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,51200,0.7546568976508247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,16384,0.24384800593058267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,65536,1.0010986328125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,16384,0.7417404386732313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,12288,0.2131697734196981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,12288,0.6383093198140463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,51200,2.1072542402479386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,10240,0.3755013412899441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,12288,0.5741173426310221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,10240,0.18067467212677002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,8192,0.3429120116763645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,8192,0.12219199869367812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,65536,2.7819716135660806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,10240,0.5200239817301432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,7168,0.30810756153530544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,7168,0.10957777500152588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,8192,0.37291733423868817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,6144,0.28583910730150014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,51200,2.8867164187961154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,6144,0.09806666771570842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,7168,0.3132462236616346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,5120,0.24026311768425834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,4096,0.16925333605872261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,6144,0.2632568942175971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,5120,0.09112177954779731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,5120,0.22531289524502227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,4096,0.06826577583948772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,4096,0.18321688969930014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,3072,0.11731378237406413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,3584,0.1509528954823812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,2560,0.09867733054690891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,3072,0.05403911074002584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,3072,0.13917244805230036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,2048,0.08652533425225152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,2560,0.048235555489857994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,2560,0.11317155096266006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,2048,0.042703999413384326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,2048,0.09796888960732354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,1024,0.045732445187038846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,1536,0.0340008901225196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,1536,0.07469689183764987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,768,0.03540800015131632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,1024,0.027826666831970215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,1024,0.05704266495174832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,3584,0.14583644602033827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,512,0.025501334004931982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,768,0.025417778227064345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,3584,0.062126219272613525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,256,0.019325332509146798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,512,0.023295111126369897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,128,0.016757332616382174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,512,0.037802666425704956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,256,0.02091022166940901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,65536,3.932584974500868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,64,0.014131555954615274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,32,0.014489778214030795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,256,0.0343466666009691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,128,0.021000888612535264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,128,0.03256533212131924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,1536,0.0672631131278144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,64,0.03261422117551168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,32,0.03241511185963949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,16384,0.44094665845235187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,768,0.04536622100406223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,65536,0.7351173294915093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,51200,0.5792773564656576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,16384,0.18111466036902535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,16384,0.5409804450141059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,12288,0.13896800412072077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,12288,0.48038313123914933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,51200,1.7240595287746852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,10240,0.27870223257276744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,12288,0.4473546875847711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,10240,0.14651555485195583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,8192,0.2692897849612766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,10240,0.3318542109595405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,8192,0.09301777680714925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,7168,0.28640355004204643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,51200,2.116355472140842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,8192,0.267247994740804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,6144,0.2160408894220988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,7168,0.26503200001186794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,6144,0.07363644573423597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,6144,0.22535199589199492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,65536,2.878150304158529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,5120,0.06402577956517537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,5120,0.1800533268186781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,3584,0.11955732769436306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,4096,0.0596017771297031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,65536,1.9921838972303603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,3072,0.10055289003584121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,3584,0.04752622379197014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,2560,0.08602399958504571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,3584,0.11701956060197617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,3072,0.04415466719203525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,2048,0.06533422072728474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,2560,0.0374017788304223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,3072,0.10207288795047337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,2560,0.08789155880610149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,1536,0.05054577853944567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,2048,0.03218488891919454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,1024,0.03426222337616815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,5120,0.14688532882266575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,2048,0.07360621955659655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,1536,0.026372444298532274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,768,0.027127110295825537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,1536,0.05797511339187622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,4096,0.12037511666615804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,512,0.02016799979739719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,1024,0.022256000174416438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,768,0.020232000284724765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,1024,0.04438222116894192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,768,0.03622044457329644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,512,0.018346667289733887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,128,0.013007111019558378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,256,0.016754666964213055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,512,0.0295466681321462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,64,0.011184888581434885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,4096,0.13492977619171143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,32,0.011237333218256632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,128,0.016770665844281513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,128,0.02507200009293026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,64,0.025056888659795124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,7168,0.08513244655397202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,256,0.015830222103330824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,256,0.026127111580636766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,65536,0.6409671041700575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,65536,1.6535511016845703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,32,0.025063110722435847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,51200,1.2065359751383464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,16384,0.47799020343356663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,16384,0.1500426663292779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,12288,0.3191111087799072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,51200,0.5001546541849772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,12288,0.11507733662923177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,10240,0.25314309861924916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,16384,0.5100933180914985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,12288,0.48324886957804364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,10240,0.3239297866821289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,8192,0.19494577248891196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,65536,2.639720916748047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,51200,1.6883849038018122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,8192,0.0797520014974806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,8192,0.22688712014092338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,6144,0.14234933588239881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,7168,0.07860977782143487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,5120,0.14419021871354845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,6144,0.06378933456208971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,7168,0.20185421572791207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,4096,0.10972622368070815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,5120,0.05440355671776665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,6144,0.16944532924228242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,3584,0.09578488932715522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,4096,0.04503822326660156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,5120,0.1405057774649726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,3072,0.0795324444770813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,4096,0.11368355486128066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,3584,0.039988444911109075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,2560,0.06383022334840563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,3072,0.03559289044804043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,2048,0.04938311047024197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,3072,0.08752089076571995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,2560,0.0325564444065094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,1536,0.03983199927541945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,10240,0.09795555803510879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,2560,0.075382219420539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,1024,0.03073244293530782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,2048,0.06197422080569797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,1536,0.02316355539692773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,768,0.020908445119857788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,1536,0.049923555718527905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,1024,0.019840889506869845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,512,0.01651733285850949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,1024,0.03831733266512553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,768,0.017665778597195942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,256,0.012413333687517377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,768,0.031543999910354614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,128,0.010104888843165504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,512,0.025581333372328017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,256,0.014850666125615438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,64,0.008958222137557136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,256,0.022487999664412603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,32,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,128,0.014890667464998035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,128,0.021672889590263367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,64,0.021690666675567627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,3584,0.10150044494205052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,32,0.021662222014533147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,2048,0.029501332177056208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,7168,0.202583114306132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,512,0.016249777542220224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,65536,0.5225804646809896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,51200,0.4064488940768772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,16384,0.3561191029018826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,51200,0.9856986999511719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,65536,1.294925371805827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,16384,0.13465332984924316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,12288,0.25728710492451984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,12288,0.09782933526568943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,10240,0.21096266640557182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,16384,0.35662221908569336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,12288,0.2715226544274224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,10240,0.09330044190088908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,8192,0.17926222748226592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,7168,0.16608177291022405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,8192,0.0663999981350369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,7168,0.06712977753745185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,6144,0.14291289117601183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,7168,0.17959377500745985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,5120,0.11410755581325954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,6144,0.06102844741609362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,6144,0.13562044832441542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,4096,0.0901528861787584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,5120,0.050260444482167564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,3584,0.08263022369808622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,5120,0.11876977814568414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,4096,0.03826755616399977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,3072,0.06708621978759766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,4096,0.09265777799818252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,3584,0.0336497790283627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,2560,0.061562663979000516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,3584,0.08776977989408706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,3072,0.02966133422321743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,3072,0.07255466779073079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,2048,0.04379466507169935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,10240,0.24255643950568306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,2560,0.02732444471783108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,1536,0.03502044412824843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,65536,1.8436205122205946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,2560,0.06336355871624418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,1024,0.024128889044125874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,2048,0.02374222212367588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,1536,0.020118221640586853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,768,0.017775111728244357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,1024,0.01686044368478987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,1536,0.04169422056939867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,1024,0.03176711002985636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,512,0.013592888911565145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,768,0.015412444869677225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,256,0.01034044474363327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,768,0.026610665851169165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,512,0.01408088869518704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,512,0.021228444245126512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,128,0.00923999978436364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,256,0.013080000049538083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,256,0.018911111685964797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,64,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,128,0.013229333692126803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,32,0.008172444171375698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,128,0.018263111511866253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,51200,1.3782115512424047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,64,0.01852799952030182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,8192,0.19138488504621717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,32,0.018360889620251127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,2048,0.05152800016933017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,16384,0.32097866800096303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,65536,0.4789511362711589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,51200,0.3836177719963922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,51200,0.8209333419799805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,65536,1.1554639604356554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,16384,0.11640889114803737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,12288,0.2331706682840983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,12288,0.08933777941597833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,10240,0.17789777119954428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,10240,0.07232355409198336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,8192,0.1589626603656345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,12288,0.27841777271694607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,51200,1.1994240019056532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,10240,0.21507733398013643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,7168,0.12119821707407634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,65536,1.6837867101033528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,6144,0.10345244407653809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,7168,0.05352888835801018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,5120,0.08596266640557183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,6144,0.05345244540108574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,7168,0.14441600110795763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,6144,0.12736799981858996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,5120,0.039113776551352605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,5120,0.10672977897855972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,3584,0.06379022200902303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,4096,0.03308444552951389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,4096,0.0838773316807217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,3072,0.06561777989069621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,3584,0.03036622206370036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,2560,0.04951644606060452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,3072,0.027229333917299908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,16384,0.36732178264194065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,2048,0.03808444407251146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,3072,0.06465066803826226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,2560,0.024792000651359558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,1536,0.02899555696381463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,8192,0.05872978104485405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,2048,0.021864000293943617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,2560,0.055983112918006055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,1024,0.020107555720541213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,2048,0.04634222057130602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,8192,0.18696622053782144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,768,0.016348444753222995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,1536,0.01850844422976176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,512,0.013111111190583972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,1536,0.037435554795795016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,768,0.014367999302016364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,1024,0.02887822190920512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,4096,0.07832533121109009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,768,0.02409244411521488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,128,0.008064000142945183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,512,0.019305777218606737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,256,0.012315555579132505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,64,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,256,0.017550221747822232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,128,0.01239377756913503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,32,0.007416889071464539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,128,0.01684533390733931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,3584,0.07416355609893799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,64,0.01686755485004849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,32,0.016827555166350473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,1024,0.015626667274369132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,65536,0.3974604341718886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,256,0.009684444301658207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,512,0.013189333180586496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,16384,0.27002843221028644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,51200,0.8279937638176812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,16384,0.09807999928792317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,51200,0.3203199969397651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,12288,0.2401199870639377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,16384,0.28272443347507054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,65536,1.3506782319810655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,51200,1.077353795369466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,10240,0.2123119963539971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,12288,0.20467289288838705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,8192,0.14284711413913304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,10240,0.06370933188332452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,10240,0.17412355211046007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,7168,0.12335999806722005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,8192,0.0517084432972802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,6144,0.100454224480523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,7168,0.045716444651285805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,8192,0.20792888270484078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,5120,0.09215733077791001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,7168,0.12719555695851645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,65536,0.9489786359998914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,6144,0.048039999273088246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,4096,0.06248444318771362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,6144,0.10628266466988458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,5120,0.03503111004829407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,3584,0.0603697763548957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,5120,0.10248977608150905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,4096,0.03435644507408142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,3072,0.04634133312437269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,3584,0.026678222748968337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,2560,0.038915554682413735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,4096,0.07278755638334486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,3072,0.025449777642885845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,3584,0.06496177779303657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,2048,0.03543555405404832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,1536,0.02693333393997616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,12288,0.07906666729185316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,3072,0.058607108063168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,1024,0.0192239996459749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,2048,0.01942755613062117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,2560,0.04957688848177592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,1536,0.016806221670574613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,2048,0.04251199960708618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,1024,0.014277334014574686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,1536,0.033752000994152494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,512,0.011735111474990845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,1024,0.02573066618707445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,768,0.013431111143694984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,256,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,768,0.021703999903466966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,512,0.012151111331250934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,128,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,512,0.017503110898865592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,256,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,128,0.011419555379284753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,32,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,128,0.014956444501876831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,64,0.01514933341079288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,32,0.015018666783968607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,2560,0.02251200046804216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,768,0.015066666735543145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,65536,0.36464622285630965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,65536,0.8039617538452148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,51200,0.6548542446560329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,51200,0.29124800364176434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,64,0.006222222414281633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,16384,0.23434755537245008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,256,0.01568266749382019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,16384,0.08693955341974895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,10240,0.14529421594407824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,51200,0.8636213408576118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,65536,1.192391077677409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,8192,0.10341155529022217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,12288,0.17525155014461943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,10240,0.054902222421434194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,7168,0.08501155508889092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,8192,0.044304887453715004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,8192,0.132950226465861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,6144,0.0830106669002109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,7168,0.03900622328122457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,7168,0.10615466700659858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,5120,0.0651128888130188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,6144,0.03633866707483927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,4096,0.05436444282531738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,6144,0.09364177783330281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,5120,0.030551999807357788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,5120,0.0783075557814704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,3584,0.05013155606057909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,4096,0.026125333375400964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,3072,0.04352088769276937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,12288,0.17954577340020073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,4096,0.06298133399751452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,3584,0.02387733260790507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,2560,0.033893333541022405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,12288,0.0646159980032179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,3584,0.056346668137444385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,16384,0.2494800090789795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,2048,0.026239110363854304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,1536,0.020266667008399963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,2560,0.02013066742155287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,3072,0.04925955666436089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,2048,0.01755911111831665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,2560,0.04310311211480034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,1024,0.015140444040298462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,2048,0.036249776681264244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,768,0.012063999970753988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,1536,0.029324445459577773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,1024,0.01310400002532535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,512,0.009339555270142024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,1024,0.022554667459593877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,768,0.012470222181744046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,256,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,768,0.019056888090239633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,512,0.011546666423479715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,512,0.015395555231306287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,256,0.011036444041464063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,256,0.013847110999955071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,64,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,128,0.01072355525361167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,32,0.005655111124118169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,128,0.013392888837390475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,3072,0.021785777476098802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,64,0.013471110827393003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,32,0.013263111313184103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,10240,0.1548817820019192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,1536,0.015463999576038785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,65536,0.561454243130154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,65536,0.29621243476867676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,16384,0.15765510665045843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,51200,0.543390221065945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,51200,0.2358462280697293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,128,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,12288,0.13677688439687094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,16384,0.07175555494096544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,16384,0.20975644058651396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,12288,0.05602400170432197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,51200,0.6855822139316134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,65536,0.9737190670437283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,12288,0.1609075599246555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,8192,0.08851022190517849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,10240,0.048093332184685596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,7168,0.07855821980370416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,8192,0.04332355658213297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,6144,0.06957599851820204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,10240,0.12621955076853433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,7168,0.03479289015134176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,5120,0.06068711148367988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,8192,0.10686666435665554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,4096,0.04861333303981357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,6144,0.032347556617524885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,7168,0.08883111344443427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,3584,0.04220889012018839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,5120,0.027394667267799377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,3072,0.03568622138765123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,5120,0.06611200173695882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,3584,0.021644444929228887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,4096,0.054995556672414146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,2560,0.02947910957866245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,3072,0.01940799918439653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,3584,0.04808355702294243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,2048,0.02292977770169576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,3072,0.04291466540760464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,2560,0.017465778523021273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,1536,0.018575110369258456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,2560,0.03736889031198289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,2048,0.015218666858143277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,1024,0.013889777991506787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,1536,0.013722666435771518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,10240,0.12021511130862766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,768,0.011113777756690979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,1536,0.0251528885629442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,512,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,768,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,1024,0.01926933394538032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,256,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,768,0.016560888952679105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,512,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,512,0.013582222163677216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,128,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,256,0.00976533359951443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,64,0.004977777600288391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,256,0.012149333126015134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,128,0.009861333502663506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,6144,0.07798133293787639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,4096,0.022848000129063923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,32,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,128,0.011751111182901593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,64,0.011768889096048145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,32,0.011775111158688864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,65536,0.2845848931206597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,2048,0.03128177920977274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,51200,0.4668133523729112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,1024,0.011757333245542316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,51200,0.2234222094217936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,12288,0.11503022246890598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,65536,0.7819822099473742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,51200,0.6230088869730631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,16384,0.17333332697550455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,12288,0.052821334865358144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,12288,0.1332471105787489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,10240,0.09513422515657212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,8192,0.07721866501702203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,10240,0.045844445625940956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,8192,0.03626044591267904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,10240,0.11600532796647813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,8192,0.09125510851542155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,6144,0.05497422152095371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,7168,0.03381866547796462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,7168,0.07962310976452298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,6144,0.029902223083708022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,5120,0.04672977659437391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,6144,0.07011644707785712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,16384,0.14884977870517307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,4096,0.036880890528361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,5120,0.025806221697065566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,16384,0.07791022459665935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,3584,0.034860445393456355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,5120,0.05981777773963081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,4096,0.022209778428077698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,3072,0.02897777822282579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,4096,0.048736890157063804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,3584,0.020285333196322124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,2560,0.024824000067181055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,3072,0.018585777944988675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,2048,0.02128355536195967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,3072,0.0391226675775316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,2560,0.016707556115256417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,1536,0.017264889346228707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,7168,0.05754399961895413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,2560,0.033886220720079206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,1024,0.013100444442696042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,2048,0.014932443698247274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,768,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,2048,0.028601778878106013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,1536,0.01346933344999949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,1024,0.011600889265537262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,1536,0.02271555529700385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,65536,0.47814755969577366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,256,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,512,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,512,0.012474666866991254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,128,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,256,0.009677333136399588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,256,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,64,0.004826666580306159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,128,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,128,0.011081777513027191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,3584,0.04382488793796963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,64,0.010968888799349466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,32,0.01092266705301073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,512,0.008613333106040955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,65536,0.5273129145304362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,768,0.010984888507260216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,65536,0.2386995686425103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,768,0.015377778146002026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,16384,0.13258755207061768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,51200,0.3404408825768365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,51200,0.1979048914379544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,12288,0.09137333101696438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,65536,0.6891839769151477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,16384,0.06557244724697538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,16384,0.1578577756881714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,10240,0.09547022316190933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,12288,0.051263110505210034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,8192,0.07074577940834893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,10240,0.04395111070738899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,12288,0.1179857783847385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,7168,0.06711200210783216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,10240,0.10217955377366807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,8192,0.034852445125579834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,6144,0.05393955442640516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,8192,0.0812657806608412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,7168,0.03137333194414774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,5120,0.04759555392795139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,7168,0.07147288984722562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,6144,0.028175999720891316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,4096,0.036489778094821505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,1024,0.0176862229903539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,6144,0.06220710939831204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,5120,0.02491377790768941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,3584,0.03258400162061056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,5120,0.053429332044389516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,4096,0.021203555994563635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,3072,0.027800000376171533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,2560,0.02461066676510705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,4096,0.043266667260064014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,3584,0.019261333677503798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,3072,0.01773244473669264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,2048,0.020206222931543987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,3584,0.03931466738382975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,2560,0.015867556134859722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,3072,0.03466755482885573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,1536,0.01572088897228241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,1024,0.011863999896579318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,2048,0.014662222729788886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,2560,0.0303182237678104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,32,0.005152889009979037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,768,0.00960355583164427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,1536,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,2048,0.0255404445860121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,512,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,1024,0.01055111073785358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,1536,0.020450666546821594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,768,0.010012444522645738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,1024,0.01626488897535536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,256,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,51200,0.6455004480150011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,128,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,768,0.01408088869518704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,64,0.0042151109211974675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,512,0.011672889192899069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,256,0.009257777697510189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,32,0.004554666578769684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,128,0.009000889129108852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,256,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,128,0.010060444474220276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,64,0.010007999837398529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,32,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,65536,0.22740001148647734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,65536,0.4414044486151801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,51200,0.3577626546223958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,51200,0.17254755232069227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,16384,0.05373866690529717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,12288,0.08585777547624375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,65536,0.5788373417324489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,51200,0.4626106686062283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,10240,0.0884560015466478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,12288,0.042372445265452065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,12288,0.10416889190673828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,8192,0.06859288613001506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,10240,0.038176000118255615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,512,0.00944800012641483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,7168,0.060271998246510826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,10240,0.0867697795232137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,8192,0.029751112063725788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,6144,0.051222221718894116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,7168,0.027835554546780054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,5120,0.04227822356753879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,7168,0.0624906685617235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,6144,0.02460177739461263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,4096,0.0318159990840488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,6144,0.054273777537875705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,5120,0.02204977803760105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,3584,0.02882844540807936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,5120,0.04602488875389099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,4096,0.01863466699918111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,16384,0.11823199854956733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,4096,0.038322667280832924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,3584,0.016943999462657504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,2560,0.021130666136741638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,3584,0.03467199868626065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,3072,0.015613334046469795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,2048,0.017457778255144756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,16384,0.14432177278730604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,2560,0.014315555493036905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,3072,0.03064889046880934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,1536,0.013824889229403602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,2560,0.02660088903374142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,1024,0.01054488867521286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,2048,0.013813333378897773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,768,0.00868711123863856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,2048,0.022414222359657288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,1536,0.011682666838169098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,1024,0.01039644413524204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,1536,0.018240888913472492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,1024,0.0147724449634552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,768,0.009707555174827576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,256,0.005033777819739448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,768,0.01260088880856832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,512,0.0092257774538464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,8192,0.07031377818849352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,128,0.00462755560874939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,512,0.010455999937322406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,64,0.004218666503826777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,256,0.008700444466537898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,32,0.00452533322903845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,128,0.008669333325492011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,256,0.009676444033781687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,128,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,64,0.009242666264375051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,3072,0.024143111374643113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,32,0.009309333231714036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,65536,0.3216968907250298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,65536,0.17134043905470106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,51200,0.2741120126512316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,51200,0.13512267006768122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,16384,0.0438435541258918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,65536,0.4976808759901259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,16384,0.12082578076256646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,10240,0.06236622068617079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,12288,0.03456622362136841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,512,0.006701333241330252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,8192,0.04963199959860908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,12288,0.09000266922844781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,7168,0.03913955556021796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,8192,0.02411199940575494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,8192,0.062239110469818115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,7168,0.021891555852360193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,7168,0.05488533443874783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,6144,0.019471999671724107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,5120,0.028209778997633193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,16384,0.08685955736372207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,12288,0.0720026691754659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,5120,0.04098755452367995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,5120,0.017557332913080852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,4096,0.02211911148495144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,51200,0.38808443811204696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,3584,0.02016266683737437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,4096,0.015578667322794596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,10240,0.03064089020093282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,4096,0.03443822264671326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,10240,0.07640710804197523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,3584,0.014170666535695394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,3584,0.031010667483011883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,2560,0.015558222929636637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,6144,0.03314666615592109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,2048,0.013182222015327878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,3072,0.027461333407296076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,2560,0.012158222496509552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,1536,0.010797333386209277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,2560,0.023589332898457844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,2048,0.011083555718262991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,1024,0.008277333445019191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,2048,0.020289777053727042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,1536,0.010030222435792288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,768,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,1536,0.016944888565275405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,1024,0.009347555538018545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,1024,0.01314933349688848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,768,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,256,0.00461422238085005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,768,0.01143555591503779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,512,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,128,0.003866666720973121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,512,0.009681777821646797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,3072,0.01812355551454756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,256,0.00872622181971868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,128,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,32,0.003556444413132138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,128,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,3072,0.013187555803192986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,64,0.008299555215570662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,32,0.008345777789751688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,65536,0.28559645016988117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,6144,0.04797955685191684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,65536,0.15478844112820095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,51200,0.22924889458550346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,16384,0.07394933038287692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,65536,0.39696889453464085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,51200,0.12371377150217693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,12288,0.05858844518661499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,16384,0.04144533475240072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,51200,0.3082035647498237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,16384,0.1026817758878072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,12288,0.03331289026472304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,256,0.008062221937709385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,12288,0.07448710997899373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,64,0.0034871109657817413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,10240,0.029672000143263075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,7168,0.03341689043574863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,10240,0.06317333380381267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,8192,0.023376888699001733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,6144,0.0273973329199685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,8192,0.05227466755443149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,7168,0.021178666088316176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,5120,0.022469333476490442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,6144,0.0188711103465822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,7168,0.047800001170900136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,4096,0.018974221414989896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,6144,0.040488001373079084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,3584,0.017383999294704862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,5120,0.034853332572513156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,4096,0.014921777778201632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,3072,0.01518311103185018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,4096,0.029166221618652344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,3584,0.014044443766276041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,512,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,2560,0.013440888788965015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,3584,0.02644266684850057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,3072,0.012886222038004132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,2048,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,3072,0.023079110516442194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,2560,0.011753777662913004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,1536,0.009290667043791877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,10240,0.048060443666246205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,1024,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,2560,0.0203813330994712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,2048,0.017501332693629794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,768,0.006361777583758037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,1536,0.014493332968817817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,1024,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,512,0.005434666656785541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,768,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,768,0.010123555858929953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,256,0.004214222232500712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,512,0.008278222547637092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,5120,0.01699822147687276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,512,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,128,0.0042408886882993905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,256,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,256,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,64,0.0037297776175869834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,128,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,32,0.004176889028814104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,128,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,64,0.007359999749395583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,65536,0.17313510841793486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,32,0.007612444460391998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,8192,0.034853332572513156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,1536,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,51200,0.15111911296844482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,65536,0.11169689231448704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,16384,0.05196977655092875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,1024,0.011759111450778114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,51200,0.08852710988786484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,65536,0.30328623453776044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,16384,0.03231111168861389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,10240,0.03101955519782172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,51200,0.23920978440178764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,12288,0.024061333802011278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,16384,0.07897244559393989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,8192,0.02584533393383026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,12288,0.05916533205244276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,10240,0.023676445086797077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,7168,0.021725333399242822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,8192,0.017229333519935608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,10240,0.05068711108631558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,6144,0.019346666004922655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,8192,0.04174044397142199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,5120,0.015930665863884818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,6144,0.015744000673294067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,6144,0.032799111472235784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,2048,0.01074044406414032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,4096,0.013736888766288757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,3584,0.012128888732857175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,5120,0.027928001350826685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,4096,0.011752888560295105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,3072,0.010832000109884473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,4096,0.02328444520632426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,3584,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,12288,0.03830666674507989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,3584,0.02129777769247691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,2560,0.009328888522254096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,2048,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,3072,0.010674667027261523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,3072,0.01908266709910499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,1536,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,2560,0.01686755485004849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,2048,0.00942755573325687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,1024,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,2048,0.014538667268223233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,1536,0.009177777502271865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,1536,0.012102222277058495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,768,0.004572444491916233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,7168,0.017238222890430026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,512,0.0038604442444112566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,1024,0.010011555420027839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,7168,0.03729333480199178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,768,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,256,0.0033848887930313745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,768,0.008750222623348236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,5120,0.014112888111008538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,512,0.0075253330998950535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,128,0.0034719999465677473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,256,0.007234666910436418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,64,0.003195555466744635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,32,0.0032595555401510666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,256,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,128,0.007276444799370236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,128,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,65536,0.1502657731374105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,64,0.00657066661450598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,32,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,51200,0.11214666896396214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,2560,0.010094222095277574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,65536,0.11089689201778835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,16384,0.04214222232500712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,51200,0.08685600095325047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,12288,0.03054222133424547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,16384,0.028976887464523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,51200,0.2054426670074463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,10240,0.0247999992635515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,12288,0.024040000306235418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,16384,0.06781511174307929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,8192,0.01977422171168857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,12288,0.05172977844874064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,10240,0.022361778550677832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,7168,0.017678222722477384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,8192,0.016794666647911072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,10240,0.04390666551060147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,6144,0.015267555912335714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,8192,0.0365226666132609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,7168,0.016562667157914903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,5120,0.013402666482660504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,512,0.007682667010360294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,4096,0.011271110839313932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,6144,0.015291555060280694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,5120,0.01312800000111262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,6144,0.028723554478751287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,3584,0.010312000082598792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,4096,0.012231999801264869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,5120,0.024745777249336243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,3072,0.009215999808576372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,4096,0.020614221692085266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,3584,0.011619555453459421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,2560,0.008358221915033128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,3584,0.01904000010755327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,3072,0.0107342220014996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,65536,0.2598035600450304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,3072,0.017101332545280457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,2560,0.010098666780524785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,1536,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,2560,0.015013333823945789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,2048,0.009688888986905416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,1024,0.005146666533417172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,1024,0.008328888979223039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,2048,0.013070222404268054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,1536,0.00923111124171151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,768,0.004378666480382283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,1536,0.010992888775136737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,1024,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,512,0.003825777934657203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,768,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,1024,0.008986666798591614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,256,0.003167111012670729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,768,0.008178667061858708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,512,0.00758844448460473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,7168,0.032582223415374756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,128,0.003195555466744635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,64,0.002862222285734283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,512,0.0070222218831380205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,256,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,128,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,128,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,65536,0.11141688293880886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,64,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,32,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,51200,0.08712622192170884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,2048,0.007078222102589077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,65536,0.10550666517681545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,16384,0.03907555672857497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,65536,0.21780089537302652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,12288,0.026903111073705886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,16384,0.028488000233968098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,51200,0.1725520028008355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,16384,0.057040002610948354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,10240,0.023392000132136877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,12288,0.022798221972253587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,8192,0.018798222144444782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,10240,0.02074044446150462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,12288,0.043947554296917386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,7168,0.01685066686736213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,10240,0.037462221251593694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,256,0.007095110913117726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,8192,0.01605955594115787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,6144,0.01516711049609714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,32,0.0031431110368834603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,5120,0.012674666941165924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,7168,0.014845333165592618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,4096,0.010785777535703448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,6144,0.013507555756303998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,7168,0.028254220883051556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,5120,0.01258488910065757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,3584,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,6144,0.024978665841950312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,3072,0.008894222478071848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,5120,0.021114667256673176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,4096,0.011100444528791638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,3584,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,3584,0.016495111915800307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,51200,0.08526933193206787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,3072,0.014917333920796713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,2048,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,2560,0.009681777821646797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,2560,0.013177778157922955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,1536,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,2048,0.00907199995385276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,1024,0.004573333180612988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,2048,0.01165600038237042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,1536,0.008642666869693333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,768,0.0038417776425679526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,1536,0.009864889085292816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,1024,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,512,0.003546666767862108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,1024,0.008455110920800103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,768,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,8192,0.031319111585617065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,256,0.0028799999919202593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,768,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,512,0.007368000017272101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,256,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,512,0.006647111227114995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,64,0.0025395556456512874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,256,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,2560,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,128,0.0069057775868309875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,3072,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,65536,0.06371644470426771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,64,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,51200,0.05304000112745497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,32,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,65536,0.10525244474411011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,51200,0.0846231116188897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,51200,0.13649511337280273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,16384,0.021589333812395733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,65536,0.1731813351313273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,12288,0.017657778329319425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,16384,0.027404444085227117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,10240,0.0159244438012441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,16384,0.0463031099902259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,12288,0.0222062220176061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,128,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,8192,0.013049778011110095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,10240,0.02029333346419864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,7168,0.011325333681371478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,12288,0.03610666592915853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,4096,0.017902221944597032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,32,0.0025600000388092464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,8192,0.02552000019285414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,8192,0.01605244477589925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,5120,0.008745777938101027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,7168,0.014927110738224454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,7168,0.023040889037979975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,4096,0.007385777930418651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,6144,0.020395555429988436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,5120,0.012464000119103326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,3584,0.006730666591061487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,4096,0.01534933348496755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,5120,0.018087110585636564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,3072,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,3584,0.010412444670995077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,3072,0.010055111514197456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,2560,0.005344888816277186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,3584,0.014335999886194864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,10240,0.031095110707812842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,2048,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,3072,0.013209777573744455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,6144,0.010006222460005019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,1536,0.004190222256713443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,2560,0.011720889144473605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,2560,0.00977955593003167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,6144,0.013735999663670858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,2048,0.010386666489972008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,1536,0.008591111335489485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,768,0.003185777821474605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,1536,0.008979555633332994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,1024,0.007928888830873702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,512,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,4096,0.011046222514576383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,1024,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,768,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,256,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,768,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,512,0.007332444190979004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,128,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,64,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,256,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,512,0.00598933340774642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,32,0.002479111154874166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,128,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,65536,0.05102488729688856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,128,0.005199111170238919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,64,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,32,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,51200,0.04048711061477661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,16384,0.015027556154463025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,65536,0.10462577475441827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,65536,0.17207554976145426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,51200,0.08425955639945136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,12288,0.012712000144852532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,16384,0.02642222245534261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,1024,0.003638222399685118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,51200,0.13573155138227674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,10240,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,2048,0.009024889104896123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,16384,0.04619644416703118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,12288,0.02198666665289137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,12288,0.035269333256615534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,7168,0.009203555683294931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,8192,0.015846222639083862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,10240,0.03013955553372701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,6144,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,8192,0.025076443950335186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,5120,0.008665777742862701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,7168,0.022544888986481562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,6144,0.013412444127930535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,4096,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,256,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,6144,0.02018311123053233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,5120,0.012132444315486483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,3584,0.006738666858938005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,5120,0.01777333352300856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,4096,0.011046222514576383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,3072,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,4096,0.0151555554734336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,2560,0.005197333378924264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,3584,0.014058666096793281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,3072,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,2048,0.004613333278232151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,3072,0.012729778057999082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,2560,0.009517333573765224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,2560,0.011628444823953839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,8192,0.009683555199040307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,1536,0.0038844446341196695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,2048,0.00904177791542477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,2048,0.010218666659461128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,10240,0.02031822171476152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,1536,0.008384000096056197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,768,0.0028693332440323303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,1024,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,1024,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,768,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,512,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,768,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,256,0.002332444406217999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,512,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,512,0.006024888820118374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,3584,0.010482666393121084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,256,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,128,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,64,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,256,0.005532444351249271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,32,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,128,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,128,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,64,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,32,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,51200,0.03778310947948032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,1024,0.003253333270549774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,16384,0.012208888928095499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,12288,0.013420444395807056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,1536,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,65536,0.17190578248765734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,10240,0.011420444481902652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,8192,0.01038755559258991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,51200,0.1354044410917494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,16384,0.045863999260796436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,7168,0.009850666754775578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,12288,0.03498488995763991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,6144,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,5120,0.007979555262459649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,10240,0.029956443442238703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,4096,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,7168,0.022595556245909795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,6144,0.020125332805845473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,3584,0.006265777680608962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,3072,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,5120,0.0177511109246148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,2560,0.004978666702906291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,4096,0.015218666858143277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,2048,0.004561777744028303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,65536,0.04692711101637947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,3584,0.0141004439857271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,1536,0.0038240001433425476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,3072,0.012788444757461548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,1024,0.003152888889114062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,768,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,2560,0.011503111157152386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,2048,0.010133333504199982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,256,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,1536,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,1024,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,7168,0.014819555812411837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,128,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,768,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,64,0.0018782222436534034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,512,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,65536,0.04731288883421156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,256,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,128,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,51200,0.03723999857902527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,64,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,16384,0.010500444306267632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,32,0.005037333402368757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,12288,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,8192,0.024902222885025874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,10240,0.008247999681366814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,8192,0.007391110890441471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,65536,0.17221332920922172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,16384,0.04590400060017904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,7168,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,6144,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,12288,0.034978667894999184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,10240,0.02982577681541443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,5120,0.010247111320495605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,8192,0.025003555748197768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,4096,0.008576889004972246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,3584,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,7168,0.022656889425383672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,512,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,6144,0.020120888948440552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,3072,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,2560,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,5120,0.017674666312005784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,2048,0.004280888785918553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,4096,0.015178667174445258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,1536,0.0038284444146686127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,1024,0.0031439999325407874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,3072,0.01276266657643848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,2560,0.011419555379284753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,768,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,2048,0.010116444693671333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,512,0.002505777817633417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,256,0.002181333386235767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,1536,0.008775110873911116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,1024,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,128,0.0019706665641731685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,64,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,768,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,512,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,256,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,128,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,64,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,32,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,3584,0.01407911048995124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,32,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,51200,0.13537155257331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,16384,0.8187928729587131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,16384,2.703103171454536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,16384,2.9127332899305554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,51200,2.9629715813530813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,12288,2.0473609500461154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,16384,2.459996541341146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,12288,0.6047822104560004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,12288,1.9977520836724176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,10240,0.5157279968261719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,10240,1.5993226369222004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,12288,1.8325494130452473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,51200,7.967705620659722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,8192,1.5230062272813585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,8192,1.3249315685696073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,10240,1.539873758951823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,10240,1.689642694261339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,8192,0.41996532016330296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,7168,1.1701013776991103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,7168,0.37167999479505753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,7168,1.192829344007704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,8192,1.1873324712117512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,6144,1.0853822496202257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,6144,0.326928006278144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,7168,1.0519413418240018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,5120,0.7748604350619845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,6144,0.9810906516181098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,5120,0.27970843844943577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,5120,0.8301839828491211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,6144,0.8592809041341146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,4096,0.6279768943786621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,4096,0.6318657663133409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,4096,0.23345777723524305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,5120,0.7232311036851672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,3584,0.538557317521837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,3584,0.20979732937282988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,4096,0.5894968774583604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,3584,0.592891534169515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,3072,0.5395004484388564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,3072,0.1837608946694268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,2560,0.34753065639071995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,3072,0.5183351304796007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,2560,0.15633066495259604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,3072,0.4525199996100531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,2560,0.6225715743170844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,2048,0.1366951068242391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,2560,0.38722578684488934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,2048,0.3382355637020535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,51200,8.17974853515625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,1536,0.25011732843187123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,2048,0.31689778963724774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,3584,0.5163760185241699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,1536,0.12735022438897028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,1536,0.27720265918307835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,768,0.11968533198038737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,1024,0.0898924469947815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,2048,0.27977334128485787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,1024,0.2097031142976549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,1536,0.24752532111273873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,51200,11.674658033582899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,512,0.08809866507848103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,1024,0.18432177437676323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,256,0.0653395586543613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,768,0.07657866345511542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,512,0.06943821907043457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,768,0.1602026621500651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,512,0.13808622625139025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,128,0.05732622411515978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,256,0.062147557735443115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,768,0.15552977720896402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,256,0.12212533420986599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,64,0.048431111706627734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,128,0.11600266562567817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,128,0.05927466683917575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,256,0.09830133120218913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,1024,0.1662622160381741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,128,0.07528533538182576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,64,0.11701511012183295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,32,0.1181066698498196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,512,0.1256435579723782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,32,0.049660444259643555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,65536,2.9670043521457248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,51200,2.332787619696723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,51200,6.093014187282986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,16384,2.319188435872396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,65536,7.993059794108073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,16384,0.6315173572964138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,16384,2.2788026597764754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,51200,9.174474080403646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,12288,1.603427569071452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,16384,1.9242729610866969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,12288,0.47080267800225156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,12288,1.6634257634480794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,10240,1.2705003950330946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,10240,0.40288978152804905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,51200,6.405016157362197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,12288,1.4373590681287978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,10240,1.4006640116373699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,65536,12.10823228624132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,8192,0.325088898340861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,8192,0.8917261759440104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,10240,1.1358782450358074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,8192,1.0996116002400715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,7168,0.9137635760837131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,7168,0.2892684406704373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,7168,0.9093484878540039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,6144,0.25418488184611004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,6144,0.7569280200534396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,8192,0.9330915874905057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,6144,0.7667786810133191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,65536,8.321714613172743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,7168,0.7764071358574761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,5120,0.5531146791246202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,5120,0.21765689055124918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,4096,0.43047200308905703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,4096,0.18948088751898873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,6144,0.6765653292338053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,5120,0.6418666839599609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,3584,0.4229413403405084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,4096,0.5331217977735732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,3584,0.16672800646887886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,4096,0.45634667078653973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,5120,0.5607022179497613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,3072,0.33317245377434623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,3584,0.46439822514851886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,3072,0.14875733852386475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,2560,0.3164088990953234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,3584,0.40159911579555935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,2560,0.1256435579723782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,2560,0.33666398790147567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,2048,0.2539920012156169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,3072,0.38556978437635636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,2048,0.1069635550181071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,1536,0.18450755543178984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,2560,0.3041155603196886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,2048,0.30508266554938424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,1024,0.12826666567060682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,1536,0.23041955629984537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,1536,0.08661333057615493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,768,0.10035200251473321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,1024,0.06894577874077691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,1024,0.16148800320095485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,1536,0.19787377781338164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,2048,0.24687200122409395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,512,0.08891910976833767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,768,0.062308443917168506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,1024,0.15247821807861328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,768,0.12669421566857234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,256,0.04907911022504171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,512,0.055323554409874805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,512,0.10925510856840347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,256,0.05050844285223219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,128,0.04532000091340807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,256,0.09710133075714111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,512,0.10105955600738525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,64,0.0377128885851966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,128,0.04827377862400479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,3072,0.3583306736416287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,128,0.09176799986097549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,128,0.061012446880340576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,32,0.09374133083555435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,768,0.12405333254072402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,256,0.07970933119455974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,64,0.09301155805587769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,65536,0.9017004436916776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,32,0.038894222842322454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,65536,2.7743235694037547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,51200,0.7755546569824219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,51200,2.2826578352186417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,16384,0.6832737922668457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,16384,0.21181600623660615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,65536,3.6219893561469187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,16384,0.7185537550184461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,12288,0.5687066713968912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,12288,0.16460355122884116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,51200,2.9948560926649304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,10240,0.42623644404941136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,12288,0.5326302316453722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,12288,0.43885689311557347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,65536,2.6899119483100047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,10240,0.1563617785771688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,16384,0.7963564660814074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,10240,0.4216177728441026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,8192,0.31191733148362905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,51200,2.0756763882107205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,8192,0.11879732873704697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,10240,0.364687098397149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,8192,0.3386133246951633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,6144,0.2263644536336263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,7168,0.0995875530772739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,8192,0.29857510990566677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,5120,0.19898311297098795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,7168,0.2689137723710802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,6144,0.25272978676689994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,5120,0.07689866754743788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,6144,0.22859644889831543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,5120,0.23710044225056967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,4096,0.14103110631306967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,5120,0.19155644045935738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,7168,0.28859999444749623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,4096,0.06398577822579278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,4096,0.17179644107818604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,3584,0.1462328831354777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,7168,0.3180239995320638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,3072,0.11593955092959934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,4096,0.15978488657209608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,3584,0.057322667704688184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,3584,0.14711288611094156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,2560,0.09651022487216526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,3072,0.050557331906424634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,6144,0.08818311161465114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,3072,0.12940177652570936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,2048,0.07969955603281657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,3584,0.14076799816555446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,2560,0.045040001471837364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,2560,0.11098666985829671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,1536,0.06362399790022108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,2048,0.03912711143493652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,3072,0.12488088342878555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,2048,0.09136711226569282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,1024,0.04312710960706075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,2560,0.10696978039211696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,1536,0.03240711159176297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,768,0.035641776190863714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,2048,0.09218400054507786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,1024,0.026679111851586237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,1024,0.05573866764704386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,512,0.02490666674243079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,1536,0.073853333791097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,768,0.023844444089465674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,768,0.04492444462246365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,1024,0.057757331265343555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,512,0.021965333157115515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,512,0.03703733285268148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,768,0.0490862230459849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,256,0.01978133287694719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,256,0.033081776565975614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,128,0.019923556182119582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,256,0.03346133232116699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,64,0.013745778136783175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,32,0.014126222994592456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,128,0.025693333811230127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,64,0.031925333870781794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,1536,0.07314133644104004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,32,0.03151911165979173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,256,0.01758311192194621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,128,0.016286222471131217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,512,0.04144444399409824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,128,0.03189155459403992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,65536,0.7169457541571723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,51200,0.561438242594401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,16384,0.6603431171841091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,65536,2.2333261701795792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,16384,0.17503644360436332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,16384,0.5586364534166124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,51200,2.236858579847548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,65536,2.8767333560519748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,16384,0.4727164374457465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,51200,1.5473599963717992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,12288,0.13710489537980822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,10240,0.3216888904571533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,12288,0.37791289223565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,51200,1.6497093836466472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,65536,2.082838270399306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,10240,0.10923288928137885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,12288,0.35025066799587673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,10240,0.31840088632371694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,8192,0.27574578921000165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,7168,0.23856533898247612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,10240,0.29157688882615834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,6144,0.19938400056627062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,8192,0.24119199646843803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,12288,0.4128862222035726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,7168,0.222908443874783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,5120,0.1383262210422092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,6144,0.06958755519655016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,7168,0.20791644520229766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,6144,0.19938133822547066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,4096,0.130895111295912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,6144,0.18002666367424858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,5120,0.05950578053792318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,5120,0.16175822416941324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,8192,0.2840329011281331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,3584,0.09818133380677964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,4096,0.05000444584422641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,5120,0.15229066212972006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,8192,0.08889155255423652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,4096,0.12827111615075007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,3584,0.04498044318623013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,3072,0.1035768919520908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,4096,0.12632889217800564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,7168,0.08124888605541654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,2560,0.07726311021380954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,3072,0.04041599896219041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,3072,0.1013066636191474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,3584,0.1136053270763821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,2048,0.06271821922726102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,2560,0.03598577777544657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,2560,0.08615289131800334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,3072,0.09906311167611016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,1536,0.050831112596723765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,2048,0.03051911128891839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,2048,0.07131910986370511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,1024,0.03313422203063965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,2560,0.08706400129530166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,1536,0.024995555480321247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,1536,0.05676355626848009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,2048,0.07338311274846394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,768,0.027221333649423387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,1024,0.021073778470357258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,1024,0.04332888788647122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,1536,0.060566220018598765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,512,0.020592000749376085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,768,0.019121777680185106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,1024,0.047998223039839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,256,0.015282667345470853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,3584,0.11401155259874131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,768,0.04066666629579332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,512,0.028471999698215063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,128,0.012406222522258759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,256,0.016148444679048326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,512,0.03454666667514377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,256,0.025547555751270715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,64,0.01091644416252772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,128,0.024712888730896845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,128,0.016004444824324716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,32,0.011340444286664328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,256,0.027653333213594224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,64,0.024640000528759424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,128,0.021359999974568684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,32,0.024703999360402424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,768,0.03528177738189697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,512,0.01737244427204132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,65536,0.6018808682759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,51200,1.2286995781792533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,16384,0.45557154549492734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,51200,0.4727733400132921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,65536,1.770130687289768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,16384,0.16546577877468532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,12288,0.3409946759541829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,51200,1.643666691250271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,16384,0.4636755519443088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,65536,2.2878621419270835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,16384,0.4376355542076959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,12288,0.10806488990783691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,10240,0.28062576717800564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,12288,0.4460959964328342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,8192,0.2249208821190728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,51200,1.3722107145521376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,12288,0.37555376688639325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,65536,1.7852463192409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,8192,0.08492889006932576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,8192,0.2193253305223253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,7168,0.2001039981842041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,7168,0.06822311215930514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,6144,0.17791377173529732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,8192,0.20559022161695692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,7168,0.21700621975792778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,5120,0.13484266069200304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,6144,0.059382220109303795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,7168,0.1784648895263672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,6144,0.16301955117119685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,5120,0.05112444361050924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,4096,0.09372978077994452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,6144,0.16639822059207493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,5120,0.1391999986436632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,10240,0.3039706548055013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,3584,0.09603199693891738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,4096,0.04327466752794054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,5120,0.1320693360434638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,4096,0.1096808910369873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,3072,0.08326577477984957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,3584,0.037646220790015325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,10240,0.25042488839891225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,3584,0.09764355421066284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,4096,0.10992355479134454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,2560,0.07230577866236369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,3072,0.033767110771603055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,3072,0.08529688914616902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,2048,0.05811466773351034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,2560,0.030192000998391047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,3584,0.0979928904109531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,2560,0.07400444481107923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,1536,0.04234666625658671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,3072,0.08715822299321492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,2048,0.025988444685935974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,2048,0.06017333269119263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,2560,0.07639555798636542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,1536,0.021436444587177698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,1024,0.0273671117093828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,1536,0.04873866505093045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,2048,0.0644764436615838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,768,0.02097333305411869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,10240,0.09326399697197808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,1024,0.01777688827779558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,512,0.016322667400042217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,1024,0.03750577900144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,768,0.016640888320075143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,768,0.030836443106333416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,256,0.012477333347002665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,512,0.02495999965402815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,768,0.03615022367901272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,512,0.030856887499491375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,256,0.014546665880415173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,128,0.010043555663691627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,256,0.024081778195169237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,128,0.02165866725974613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,64,0.008911111288600499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,128,0.014003554979960123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,128,0.0188417782386144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,32,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,64,0.021354667014545862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,1536,0.05319911241531372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,32,0.021187555458810594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,1024,0.04119466741879781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,512,0.01512977812025282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,256,0.02223199937078688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,65536,0.497744878133138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,65536,1.4819697274102106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,16384,0.3223200109269884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,51200,1.1538320117526584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,51200,0.38927377594841855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,16384,0.1318302286995782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,51200,1.2874008814493816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,12288,0.2937386565738254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,16384,0.3930159939659967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,16384,0.3113679885864258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,10240,0.2274097866482205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,12288,0.1013137764400906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,12288,0.2740195592244466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,65536,1.3958053588867188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,51200,1.0740426381429036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,8192,0.187462223900689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,12288,0.2374168766869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,10240,0.21824799643622506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,8192,0.06355377700593737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,7168,0.1359484460618761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,10240,0.20331022474500868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,8192,0.22259200943840873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,6144,0.13496977753109402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,8192,0.17121333546108672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,7168,0.05606844690110949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,65536,1.7699413299560547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,5120,0.10655733611848618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,6144,0.04964355627695719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,6144,0.13940355512830946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,7168,0.14820444583892822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,4096,0.08522133032480876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,5120,0.04220088985231188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,6144,0.12626132700178358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,5120,0.11266667313045925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,4096,0.03572444452179803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,4096,0.09102755784988403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,10240,0.07712888717651367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,5120,0.10799200005001491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,3072,0.0684915582338969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,3584,0.030985779232449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,4096,0.09155378076765273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,2560,0.0546320014529758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,3072,0.027975112199783325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,3584,0.0817777779367235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,3072,0.07049689027998182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,7168,0.15444000562032065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,2048,0.042502221133973866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,3072,0.07330133517583211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,2560,0.061250666777292885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,2560,0.025751110580232408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,1536,0.03204800022972955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,2048,0.02233955595228407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,2560,0.06263644165462919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,2048,0.05077155431111654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,1024,0.0240666667620341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,1536,0.04078844520780776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,1536,0.019300444258583915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,2048,0.05407999952634176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,768,0.0188719994492001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,1024,0.016000888413853116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,1536,0.04533777634302775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,512,0.014243554737832812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,768,0.014476444986131458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,1024,0.034959110948774554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,768,0.026201777988009985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,3584,0.08021955357657538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,256,0.01074399964676963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,512,0.02091644373204973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,512,0.013403555585278405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,256,0.012489777472284106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,256,0.018765333626005385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,128,0.009296000003814697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,512,0.025425778494940862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,128,0.012601777911186218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,64,0.007881777981917063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,128,0.017904000149832833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,32,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,128,0.01718311177359687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,3584,0.07750578059090509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,1024,0.03136266602410211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,32,0.017887110511461895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,768,0.030745777818891738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,65536,0.4424995581309001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,256,0.02163644466135237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,65536,1.4360809326171875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,64,0.018233777748213876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,51200,0.9007600148518881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,16384,0.3268231021033393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,65536,1.6368462244669597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,51200,0.3575715488857693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,16384,0.3135840098063151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,16384,0.10846577750311957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,65536,1.26872624291314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,12288,0.27246665954589844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,12288,0.08117066489325629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,51200,1.2361733118693035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,10240,0.2227706644270155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,12288,0.2317288981543647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,10240,0.077437334590488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,51200,0.9905413521660699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,8192,0.17594756020439994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,12288,0.23138488663567436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,10240,0.2136311133702596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,7168,0.15037777688768175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,8192,0.06258755260043673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,8192,0.15528799427880183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,10240,0.1986915535397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,6144,0.12784000237782797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,7168,0.054683556159337364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,8192,0.172160890367296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,5120,0.10453777843051487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,7168,0.14335466755761042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,6144,0.04485422372817993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,6144,0.11908000045352513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,4096,0.0849813355339898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,5120,0.03769066598680284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,6144,0.12263822555541992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,5120,0.10004622406429714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,3584,0.07083200083838569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,4096,0.031947556469175555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,16384,0.33269333839416504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,5120,0.10754666725794475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,3072,0.06439644760555692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,3584,0.02936089038848877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,4096,0.08860177463955349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,3584,0.07217599948247273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,2560,0.04925422204865349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,3072,0.026891556051042344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,3072,0.06407199965582953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,3584,0.07834488815731473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,2048,0.036863111787372164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,2560,0.02421599957678053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,3072,0.06989600261052449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,7168,0.1373022264904446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,2560,0.0554017788834042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,1536,0.030159112479951646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,2048,0.02068000038464864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,2560,0.0616862244076199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,1024,0.019952888290087383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,1536,0.01760444376203749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,1536,0.036862221029069685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,2048,0.05258488986227247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,768,0.01682044400109185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,1024,0.014764444695578681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,1024,0.028607110182444256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,512,0.012934221989578672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,4096,0.08181866672303942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,768,0.013764444324705334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,1024,0.033944000800450645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,768,0.023800000548362732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,256,0.009407999614874521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,512,0.012688889271683164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,512,0.01888533267709944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,768,0.028543111350801256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,128,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,256,0.01685333251953125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,512,0.023911111884646948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,64,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,256,0.02092711130777995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,128,0.016325333052211337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,32,0.007501333124107785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,2048,0.04552355408668518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,128,0.016540444559521146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,64,0.016488000750541687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,32,0.016315556234783597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,1536,0.043043557140562266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,65536,0.3885235521528456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,256,0.012050666742854647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,51200,0.7224995295206705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,65536,1.1914613511827257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,128,0.011867555479208628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,16384,0.32413776715596515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,65536,1.3930240207248266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,51200,0.2993137836456299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,16384,0.2699813312954373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,12288,0.20013511180877686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,16384,0.09554400046666463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,65536,1.1008639865451388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,51200,1.0564781824747722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,12288,0.07279022534688313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,12288,0.20180266433291963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,16384,0.36481867896185977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,10240,0.06143199735217624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,51200,0.8628311157226562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,10240,0.18143910831875273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,8192,0.1341466638776991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,12288,0.21258755524953207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,7168,0.11965421835581462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,8192,0.05046755406591627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,10240,0.1739342212677002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,8192,0.13653777705298528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,6144,0.10050044457117717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,7168,0.12577244970533583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,7168,0.04481066597832573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,8192,0.1507964399125841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,6144,0.039693332380718656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,7168,0.1293920013639662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,6144,0.10470400253931682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,5120,0.03424266642994351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,4096,0.06553155846065946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,5120,0.0878524449136522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,6144,0.10986133416493733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,10240,0.16836178302764893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,3584,0.0533884432580736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,4096,0.02940000096956889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,5120,0.09473333093855117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,3072,0.05117244521776835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,4096,0.07189866569307116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,3584,0.026495110657480028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,2560,0.04064800010787116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,3584,0.06304088566038343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,3072,0.02420444455411699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,4096,0.0787680016623603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,2048,0.03421866562631395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,3072,0.05555999941296048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,3584,0.07164622015423246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,2560,0.022077333596017625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,1536,0.02641155487961239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,3072,0.06235022015041775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,2560,0.048468444082472056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,2048,0.019137778215938143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,1024,0.018752889500723947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,2048,0.040158222119013466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,2560,0.054788443777296275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,1536,0.016639111770523917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,768,0.015340444114473132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,1536,0.03323910964859857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,1024,0.014168888330459595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,1024,0.025244444608688354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,1536,0.03881866733233134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,512,0.01146044416560067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,768,0.013157332936922709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,1024,0.03106133474244012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,768,0.021244444780879553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,256,0.011143111520343356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,512,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,768,0.026343110534879897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,512,0.016951110627916124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,128,0.009344889058007134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,256,0.011424889167149862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,256,0.015461333923869662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,512,0.022706665926509436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,64,0.008832889298597971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,32,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,128,0.01482399966981676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,256,0.019229332605997723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,128,0.015422221687104968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,32,0.014720888601409065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,2048,0.0469084448284573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,65536,0.33792000346713597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,65536,1.020491600036621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,51200,0.7398266792297363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,65536,1.059724489847819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,128,0.01145511120557785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,64,0.01481244464715322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,16384,0.20565511120690239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,5120,0.08903021944893731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,51200,0.28020532925923664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,16384,0.22613154517279732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,65536,0.9571395450168185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,16384,0.08041777875688341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,51200,0.8280408647325305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,10240,0.13807200060950384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,12288,0.06609511375427246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,16384,0.23501245180765787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,12288,0.18325777848561606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,10240,0.05548444390296936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,8192,0.1178462240431044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,10240,0.15451110733879939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,12288,0.17788533369700113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,7168,0.09594577550888062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,51200,0.7537822193569608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,8192,0.04315288861592611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,10240,0.1517546706729465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,8192,0.11937688456641303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,6144,0.07739644580417208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,7168,0.03849955399831136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,8192,0.12514221668243408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,7168,0.10294310914145575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,5120,0.06809777683681913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,6144,0.03447466757562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,12288,0.18123466438717314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,7168,0.11927733156416152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,4096,0.057743999693128795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,5120,0.030066665675905015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,6144,0.0963111122449239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,3584,0.04762311114205254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,4096,0.06487466891606648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,4096,0.025592888394991558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,5120,0.08379199769761826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,3072,0.0409804450141059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,3584,0.02356622285313076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,3584,0.05559111303753323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,3072,0.021752888957659405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,3072,0.04923911227120293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,2560,0.03303555647532145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,3072,0.0553288890255822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,2560,0.042861332496007286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,2048,0.025225778420766194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,2560,0.019641778535313077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,2560,0.048920889695485435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,5120,0.07565422190560235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,1536,0.0214044451713562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,2048,0.017440888616773818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,1536,0.028912001185946997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,4096,0.07067733340793185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,1024,0.015722667177518208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,1536,0.01475644442770216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,1536,0.034930666287740074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,3584,0.061800890498691134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,768,0.012251555919647217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,1024,0.012716444830099741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,768,0.01200355589389801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,1024,0.027085334062576294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,768,0.018718222777048748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,6144,0.09082844522264268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,768,0.02400622268517812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,512,0.011071999867757162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,256,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,512,0.01517511076397366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,2048,0.03608533408906724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,128,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,512,0.02001688877741496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,256,0.010407999985747868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,256,0.013775111072593264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,64,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,2048,0.04188266727659437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,256,0.01789688898457421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,32,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,128,0.013129777378506131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,64,0.013190222283204397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,128,0.01421511173248291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,1024,0.02260355485810174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,32,0.013083555632167392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,512,0.009390222529570261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,65536,0.577397346496582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,65536,0.2932284408145481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,51200,0.5212000211079916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,16384,0.18265333440568712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,51200,0.23015467325846353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,65536,0.8794719907972547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,51200,0.6862710846794976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,128,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,12288,0.12968622313605413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,16384,0.20550755659739176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,16384,0.06898133622275458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,10240,0.10843555794821845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,65536,0.8286115328470866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,12288,0.054304000404145986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,51200,0.610063976711697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,16384,0.2109608915117052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,12288,0.14358222484588623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,8192,0.09332533015145196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,10240,0.05334933267699348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,10240,0.12098756101396348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,8192,0.03651644455062018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,8192,0.11010399791929458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,10240,0.13265600469377306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,7168,0.03338933322164748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,6144,0.06753244664933947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,7168,0.09472444322374131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,8192,0.10684888892703587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,5120,0.05711911122004191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,6144,0.030015110969543457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,7168,0.09421066443125407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,6144,0.07460444503360324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,5120,0.06394400199254353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,5120,0.026346666945351496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,6144,0.08318399720721774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,5120,0.07270755370457967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,4096,0.02250755495495266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,3584,0.03863022062513564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,4096,0.05185866687032911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,7168,0.08247733116149902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,3584,0.020397333635224234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,12288,0.15420356061723497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,3584,0.04661155409283108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,3584,0.055213332176208496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,3072,0.018925334016482036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,2560,0.028723554478751287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,3072,0.04938222302330864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,4096,0.04543733265664843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,2560,0.03638844357596503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,2048,0.024086222052574158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,2560,0.01703733371363746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,2560,0.04325955443912082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,1536,0.01902577777703603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,2048,0.015174221661355762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,2048,0.0303457776705424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,1024,0.013793778088357715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,1536,0.01349511080318027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,2048,0.03644444545110067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,1536,0.024412444896168176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,4096,0.060396446122063525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,768,0.011351111034552256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,3072,0.033542222446865506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,1024,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,1024,0.018748443987634447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,1536,0.030464887619018555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,512,0.008811555802822113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,768,0.011056000159846412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,256,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,768,0.016229333149062265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,3072,0.04126844472355313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,512,0.010281778044170803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,512,0.013125333521101209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,768,0.021309332715140447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,128,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,256,0.009860444400045607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,64,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,256,0.012120000190205045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,512,0.018564444449212816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,32,0.006463999963468975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,128,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,256,0.015726221932305228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,128,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,64,0.011434666812419891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,128,0.012992888689041138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,32,0.011382222175598145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,65536,0.27004265785217285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,65536,0.5784382290310329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,51200,0.5064969062805176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,16384,0.1640062199698554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,51200,0.21397599909040663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,65536,0.774993790520562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,51200,0.612486203511556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,12288,0.12353066603342693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,16384,0.06698755423227946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,16384,0.19288267029656303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,65536,0.7944275538126627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,51200,0.6182595358954536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,12288,0.05175288849406772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,16384,0.20015645027160645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,12288,0.1288871102862888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,8192,0.08633689085642497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,10240,0.044525331921047635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,10240,0.11264888445536296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,12288,0.15221688482496473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,7168,0.07648177941640218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,8192,0.08879022465811835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,10240,0.13082133399115667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,6144,0.06359733475579156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,8192,0.106385780705346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,7168,0.07853688796361287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,1024,0.024663999676704407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,5120,0.051242666112052075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,6144,0.02840177880393134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,7168,0.09556621975368923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,6144,0.06867021984524198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,4096,0.04222311245070564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,10240,0.1093555556403266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,6144,0.08245155546400282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,5120,0.05825510952207777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,3584,0.037399109866884016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,4096,0.021532444490326777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,4096,0.04864711231655545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,5120,0.07033066617117988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,3072,0.033075554503334895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,3584,0.01995822290579478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,3584,0.04348266786999173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,4096,0.05856088797251383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,8192,0.035303112533357404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,2560,0.028063999281989202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,3072,0.018253333038753934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,3072,0.03843733337190416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,3584,0.05361955695682102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,7168,0.03193777799606323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,2560,0.01683911184469859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,2560,0.03371289041307237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,3072,0.04823911190032959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,1536,0.019075555933846366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,2560,0.04156800111134847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,2048,0.014675555957688225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,2048,0.02786933382352193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,1024,0.0144533332851198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,5120,0.024862221545643274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,1536,0.01312622262371911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,1536,0.022815111610624526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,2048,0.03560444381501939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,768,0.011076444553004371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,1024,0.011542222566074796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,1024,0.01756533318095737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,1536,0.029558221499125164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,512,0.008820444345474243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,768,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,768,0.015139554937680563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,256,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,512,0.010055111514197456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,512,0.012404444317022959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,768,0.020571556356218126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,512,0.017649778061442904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,256,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,256,0.011198221809334226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,2048,0.02363288899262746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,64,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,32,0.0063075555695427795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,128,0.009653333160612319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,128,0.010674667027261523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,256,0.015485333071814643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,64,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,32,0.010690666735172272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,128,0.012755555411179861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,65536,0.5412524541219076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,65536,0.22598934173583984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,1024,0.02350666622320811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,16384,0.1367617713080512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,51200,0.18242043919033477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,51200,0.5307200219896104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,65536,0.6406906445821127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,12288,0.09944088591469659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,16384,0.060079998440212674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,16384,0.1495751142501831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,12288,0.04681422313054403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,10240,0.0982159972190857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,51200,0.5119386778937446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,12288,0.11923821767171223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,16384,0.1761031150817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,8192,0.07882222202089098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,10240,0.0425911115275489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,12288,0.12690756056043836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,10240,0.09789155589209662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,51200,0.33881600697835285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,8192,0.03216800093650818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,8192,0.0792328914006551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,10240,0.10734488566716512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,6144,0.05611199802822537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,7168,0.029446221060223047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,65536,0.6697440147399902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,7168,0.07109422153896756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,8192,0.08957688675986396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,5120,0.04575111137496101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,6144,0.025216889050271776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,4096,0.03789155681927999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,7168,0.07916444540023804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,5120,0.02288888891537984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,5120,0.052984890010621816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,6144,0.0695448915163676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,4096,0.019632889164818656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,4096,0.043339557117886014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,5120,0.060533331500159375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,3072,0.027826666831970215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,4096,0.05061155557632446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,3584,0.01757333344883389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,3584,0.03897688786188761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,2560,0.023170666562186346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,3072,0.015989333391189575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,7168,0.06426578097873263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,3072,0.03448977735307481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,2048,0.019360888335439894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,2560,0.014690667390823364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,3072,0.04110666778352525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,2048,0.02496799992190467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,1536,0.013387555877367655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,2048,0.013703111145231458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,2560,0.035924444595972695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,6144,0.06081333425309923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,1024,0.010288889209429422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,1536,0.011993777420785693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,2048,0.0313297775056627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,3584,0.03128000100453695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,1536,0.020367999871571858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,768,0.008310221963458592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,1024,0.010415111151006486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,1024,0.015981333123313058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,512,0.007034666836261749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,1536,0.025748444928063288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,768,0.009995555712117089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,768,0.013879111243618859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,1024,0.02058844433890449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,256,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,768,0.018852444158660043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,512,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,3584,0.04611733224656847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,128,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,256,0.008735111190213097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,256,0.010328888893127441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,2560,0.030159112479951646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,512,0.017553778158293832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,64,0.004699555536111196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,32,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,128,0.008913777768611908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,128,0.009716444545321995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,256,0.0148044443792767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,64,0.009728000395827824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,32,0.00977600034740236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,65536,0.4573857519361708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,65536,0.20766044987572563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,51200,0.32083378897772896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,65536,0.5718462202284071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,512,0.011415111521879831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,16384,0.11618399620056152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,51200,0.16493954923417833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,65536,0.6268826590643989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,51200,0.4899626837836371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,16384,0.05335999859703911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,12288,0.08753422233793472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,16384,0.13351199362013075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,10240,0.057167112827301025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,12288,0.044216887818442456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,51200,0.490772459242079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,128,0.011704888608720569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,12288,0.10670222176445855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,8192,0.054076443115870156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,10240,0.03700977894994948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,10240,0.08530844582451715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,12288,0.12568088372548422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,7168,0.04122666517893473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,8192,0.0292595558696323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,8192,0.06974133518007067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,6144,0.04331022169854906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,10240,0.10705244541168213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,7168,0.02649333410792881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,5120,0.0318213336997562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,7168,0.06133244434992472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,8192,0.08790044652091132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,6144,0.02390488816632165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,6144,0.053824888335333936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,4096,0.029280887709723577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,7168,0.07736177576912774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,5120,0.020978666014141504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,3584,0.025923555095990498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,5120,0.045943111181259155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,6144,0.06914489136801825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,4096,0.017996443642510306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,3072,0.02132266594303979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,5120,0.059567113717397056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,3584,0.016789333687888253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,3584,0.034487111700905695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,16384,0.1654613282945421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,2560,0.01721333298418257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,3072,0.01550222271018558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,3072,0.030415998564826116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,3584,0.04492355717553032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,2048,0.014286221729384528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,2560,0.014121777481502958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,2560,0.02637600070900387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,3072,0.04023822148640951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,1536,0.012065778175989786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,2048,0.013099555340078143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,2048,0.02209866709179348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,1024,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,1536,0.011764444410800934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,1536,0.018259555101394653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,2048,0.030119998587502375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,768,0.007585778004593319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,1536,0.025441777375009324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,1024,0.010130667024188572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,1024,0.014504000544548035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,512,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,4096,0.038147555457221136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,768,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,1024,0.020279111133681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,256,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,4096,0.04966577887535095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,512,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,512,0.009282666775915358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,128,0.004547555413511065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,768,0.018227555685573153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,256,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,256,0.009708444277445475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,512,0.01569688816865285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,64,0.004521777646409141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,32,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,128,0.008683555656009251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,128,0.009187555147541894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,256,0.014819555812411837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,2560,0.035559111171298556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,64,0.00905777762333552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,128,0.011055111057228513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,32,0.009033777647548253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,65536,0.29564444224039715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,51200,0.27398843235439724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,65536,0.16434666845533583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,16384,0.09092977974149917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,768,0.012542222109105853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,65536,0.47788000106811523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,51200,0.13107111718919542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,51200,0.37757688098483616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,12288,0.07049422131644355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,16384,0.04310133390956455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,16384,0.11273688740200466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,65536,0.48996978335910374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,10240,0.0561457806163364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,51200,0.38546400600009495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,12288,0.033751110235850014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,12288,0.08570488956239487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,8192,0.04609510964817471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,16384,0.131022228135003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,10240,0.030251555972629126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,10240,0.07220444414350721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,7168,0.03787022166781955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,12288,0.1009662217564053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,8192,0.023334221707450017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,8192,0.05924266576766968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,6144,0.035755554835001625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,7168,0.021686222818162706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,5120,0.030029333300060693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,7168,0.052432888083987765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,6144,0.019741333193249173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,6144,0.04594399862819248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,7168,0.06327999962700738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,4096,0.023792889383104112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,5120,0.017253332667880587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,5120,0.039447112215889826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,3584,0.021746666895018682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,4096,0.015095111396577625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,5120,0.048269334766599864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,4096,0.03275377882851495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,3072,0.018941332896550495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,4096,0.04080533319049411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,3584,0.029683556821611192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,3072,0.026160000099076167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,3072,0.013088000317414602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,2560,0.0162426663769616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,3584,0.03699288765589396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,10240,0.08422222402360703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,2560,0.012106666962305704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,2048,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,2560,0.022514666120211285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,8192,0.07155911127726237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,1536,0.011158222125636207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,2048,0.010954666468832227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,2048,0.0192586663696501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,1536,0.015882665912310284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,6144,0.05567110909356011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,2048,0.02570844524436527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,1024,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,1024,0.009180444810125563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,1536,0.021319111188252766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,3584,0.014209777116775513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,1024,0.012808000048001608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,1024,0.017511111166742113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,512,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,768,0.008964444200197855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,768,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,512,0.008353778057628209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,3072,0.03355111016167535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,256,0.004518222063779831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,512,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,2560,0.029500444730122883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,512,0.014152889450391134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,256,0.008909333083364699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,128,0.00424177779091729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,1536,0.010058666268984476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,256,0.013100444442696042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,128,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,32,0.004221333397759331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,128,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,768,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,128,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,64,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,32,0.008348444269763099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,65536,0.25871555010477704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,51200,0.18381955888536242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,65536,0.15461333592732748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,768,0.016051555673281353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,65536,0.39414045545789933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,16384,0.07694044378068712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,51200,0.3072435590955946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,51200,0.12221066157023112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,256,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,12288,0.05633599890602959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,16384,0.041195554865731135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,65536,0.4792906443277995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,16384,0.09973866409725612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,10240,0.04807733496030172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,64,0.003871110992299186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,12288,0.07456444369422065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,51200,0.3766453266143799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,16384,0.12963199615478516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,8192,0.036229332288106285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,10240,0.028880887561374243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,12288,0.09847999943627252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,10240,0.06329066885842217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,7168,0.032136887311935425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,8192,0.022732444935374792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,8192,0.051295999023649425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,10240,0.08410133255852593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,6144,0.027472888429959614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,7168,0.020619556307792664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,7168,0.04627644353442722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,8192,0.06920444303088717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,5120,0.02368800010946062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,6144,0.01865955524974399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,7168,0.06217333343293932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,6144,0.04004355602794223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,4096,0.019327110714382596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,5120,0.016587555408477783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,5120,0.03463200065824721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,6144,0.05502666698561775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,3584,0.016764443781640794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,4096,0.01482311056719886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,4096,0.02882755464977688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,5120,0.04798488815625509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,3072,0.015396444333924187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,12288,0.03199911117553711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,3584,0.013781333135233985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,4096,0.04007555709944831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,2560,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,3072,0.012792000340090858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,3072,0.02316177719169193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,3584,0.035487999518712364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,2048,0.011439111497667102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,2560,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,2560,0.020413332515292697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,3072,0.03202133377393087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,1536,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,2560,0.028747555282380845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,1536,0.014524444937705994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,2048,0.02479022244612376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,1536,0.009859555297427708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,1024,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,1536,0.021137777302000258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,1024,0.00905688852071762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,768,0.005877333382765452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,1024,0.011582222249772815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,1024,0.01711466742886437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,768,0.008692444198661381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,768,0.010081777969996134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,512,0.005163555757866966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,3584,0.025834666358100042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,256,0.0040755557517210645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,768,0.01609244445959727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,512,0.008328888979223039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,512,0.00868977771864997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,128,0.0035537779331207275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,256,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,256,0.007719111111429002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,64,0.0032071111102898917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,2048,0.01092977739042706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,128,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,32,0.0034862222770849862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,256,0.01276533305644989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,128,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,64,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,65536,0.18622667259640166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,128,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,32,0.0075706665714581805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,51200,0.13836356004079184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,65536,0.10754755470487808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,65536,0.3011733425988091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,16384,0.051903112067116626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,51200,0.08681688706080119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,51200,0.23544179068671334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,65536,0.34090044763353133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,16384,0.029143999020258587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,16384,0.07846133576499091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,51200,0.26705066363016766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,10240,0.03024266825781928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,16384,0.09511555565728082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,512,0.014144889182514615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,12288,0.05932266844643486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,2048,0.017512000269360013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,8192,0.02471911079353756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,10240,0.021321778496106465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,10240,0.05033333433998955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,12288,0.07276622454325359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,7168,0.02208799951606327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,8192,0.016626665989557903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,8192,0.04141777753829956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,10240,0.06229866875542534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,6144,0.018419555491871305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,7168,0.015339555011855232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,8192,0.052411556243896484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,7168,0.03720888826582167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,5120,0.016009777784347534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,6144,0.013814222481515674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,7168,0.046353777249654136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,12288,0.03823555509249369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,4096,0.013841778039932251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,5120,0.02796799937884013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,5120,0.012621333201726278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,4096,0.011363555159833698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,12288,0.023385778069496155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,4096,0.023295111126369897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,4096,0.0303422212600708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,3584,0.021205332544114854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,3072,0.011003555523024665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,3584,0.010845333337783813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,3584,0.027679999669392902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,3072,0.018936889039145578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,3072,0.010416000253624385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,2560,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,6144,0.032615999380747475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,3072,0.02498666610982683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,2048,0.008053333395057255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,6144,0.041011555327309504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,2560,0.01664533383316464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,2560,0.02223644488387638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,2048,0.009354666703277165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,3584,0.012497777740160624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,1536,0.006776888751321369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,2048,0.014490667316648694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,1024,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,1536,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,2048,0.01995644470055898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,1536,0.012081777883900536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,768,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,1024,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,1024,0.008126222425036961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,1536,0.017217778497272067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,512,0.00397866674595409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,768,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,768,0.008737777670224508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,256,0.003559111307064692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,512,0.007358222372002072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,768,0.013380444712109037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,2560,0.009712888962692684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,128,0.0028942223224374983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,256,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,512,0.012375999655988483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,256,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,64,0.0028284444577164124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,5120,0.03564711080657111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,32,0.0028115556471877625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,128,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,256,0.011394666300879585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,65536,0.16601067119174534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,64,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,32,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,51200,0.1299440066019694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,65536,0.25782044728597003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,1024,0.014152000347773233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,512,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,51200,0.08484177456961738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,16384,0.04191822144720289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,65536,0.3371315532260471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,16384,0.028765334023369685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,51200,0.26547198825412327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,128,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,12288,0.029895110262764826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,16384,0.06787200106514825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,10240,0.024255999260478552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,12288,0.023200000325838726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,128,0.009014222357008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,12288,0.051190220647388034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,8192,0.018998222218619455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,10240,0.04371733466784159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,10240,0.020968889196713764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,12288,0.07147910859849718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,65536,0.10548177692625259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,7168,0.01753155555990007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,8192,0.01645599967903561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,10240,0.0611048870616489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,6144,0.015757333901193406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,7168,0.015172445111804538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,7168,0.0325564444065094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,8192,0.05186577637990316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,51200,0.20441688431633842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,5120,0.013425778183672162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,7168,0.046022223101721875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,6144,0.028743998871909246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,6144,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,4096,0.011400000088744693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,5120,0.012497777740160624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,5120,0.02443555494149526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,3584,0.010308444499969482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,6144,0.040638221634758845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,4096,0.011369778050316704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,4096,0.020577778418858845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,16384,0.0930817789501614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,3072,0.009153777526484596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,3584,0.010671111444632212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,4096,0.029091556866963703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,2560,0.008230222596062554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,3072,0.017055999901559617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,3072,0.01039644413524204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,3584,0.026745777991082933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,8192,0.03606399893760681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,3072,0.02457422183619605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,2560,0.010079999764760336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,2560,0.015136000182893542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,2560,0.021947556071811255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,1536,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,2048,0.013071111506885953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,2048,0.019367999500698514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,1536,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,1024,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,1536,0.008836444053384993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,5120,0.03495555453830295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,1536,0.01685688893000285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,3584,0.018868444694413077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,768,0.00453599997692638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,1024,0.00904088881280687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,1024,0.014127999544143677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,768,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,512,0.003724444243643019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,2048,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,768,0.013110222087966071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,512,0.007363555332024892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,256,0.003555555724435382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,512,0.0069884442620807225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,2048,0.009307555854320526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,128,0.003212444484233856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,512,0.011810666984981961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,256,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,128,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,128,0.007003555695215861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,64,0.003006222140457895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,32,0.0030462222380770575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,1024,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,65536,0.11191999912261963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,128,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,64,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,32,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,51200,0.0835991104443868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,768,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,65536,0.104567997985416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,16384,0.03390489021937052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,65536,0.2187591128879123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,51200,0.08384533060921563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,51200,0.17220977942148843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,12288,0.02471911079353756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,16384,0.05724977784686618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,65536,0.32325243949890137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,256,0.007113777928882175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,10240,0.020344000723626878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,51200,0.2551306618584527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,12288,0.02233422299226125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,12288,0.04393777913517422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,16384,0.09038044346703424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,8192,0.01697688963678148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,256,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,10240,0.020367999871571858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,10240,0.03764355513784621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,7168,0.015340444114473132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,12288,0.06883822547064887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,8192,0.01587288909488254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,6144,0.013397333522637686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,8192,0.031180444690916274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,10240,0.059247997072007924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,7168,0.014526221487257214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,5120,0.011626666618718041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,8192,0.04955377843644884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,7168,0.028299556838141546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,6144,0.013427555561065674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,4096,0.009911999934249455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,6144,0.02494577732351091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,7168,0.044086221191618175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,5120,0.012123555772834353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,3584,0.009096889032257928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,5120,0.021356445219781663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,6144,0.03858577873971727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,4096,0.011011555790901184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,3072,0.008221333225568136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,16384,0.02752444479200575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,5120,0.03381866547796462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,3584,0.01037333326207267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,2560,0.007247111035717859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,3584,0.016536000702116225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,4096,0.028211557202868994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,3072,0.010030222435792288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,2048,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,3584,0.02597066594494714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,2560,0.009648000200589498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,2560,0.013288000391589271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,1536,0.0053048887186580235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,3072,0.023356444305843774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,2048,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,2048,0.01146933353609509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,1024,0.004280000097221798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,1536,0.009873777627944946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,1536,0.008307555483447181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,2048,0.018906666172875296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,768,0.0038693332009845306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,1024,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,1536,0.016194666425387066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,1024,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,768,0.0075911109646161394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,4096,0.01793066660563151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,512,0.0032684444967243406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,1024,0.013963555296262106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,768,0.013059555656380124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,256,0.0028115556471877625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,3072,0.014850666125615438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,512,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,512,0.011740444435013665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,128,0.002521777732504739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,256,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,2560,0.021385777327749465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,256,0.011012444065676795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,128,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,64,0.0023093333260880574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,32,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,128,0.005628444254398346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,65536,0.0661804477373759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,128,0.008611555728647444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,64,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,32,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,51200,0.052298665046691895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,768,0.00740444411834081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,65536,0.10404444403118557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,65536,0.1728515492545234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,512,0.007244444555706448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,16384,0.017817777064111497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,51200,0.0835297769970364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,51200,0.13631644513871935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,256,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,12288,0.016936888297398884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,65536,0.31523556179470485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,16384,0.0463013317849901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,16384,0.02695111102528042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,10240,0.015481778317027621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,51200,0.25155732366773814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,12288,0.021480000681347314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,16384,0.08827110793855454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,12288,0.03595644566747878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,10240,0.019949333535300363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,10240,0.030962665875752766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,7168,0.011743110915025076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,12288,0.06741866800520155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,8192,0.015561777684423657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,10240,0.05743910868962606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,8192,0.025595555702845257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,6144,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,7168,0.014452444182501899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,7168,0.0229120006163915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,5120,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,8192,0.0481697784529792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,6144,0.01348977784315745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,4096,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,6144,0.020385776956876118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,7168,0.04270844327078926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,5120,0.012063999970753988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,5120,0.01792888840039571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,3584,0.006993778049945831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,4096,0.010986666712496014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,4096,0.015384000208642749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,5120,0.03239911132388645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,3072,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,3584,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,3584,0.01420088940196567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,4096,0.027456889549891155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,2560,0.005848888721730974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,3072,0.010058666268984476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,8192,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,3584,0.02549244463443756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,2048,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,2560,0.009375111096435124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,2560,0.011668444507651858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,3072,0.023141332798533972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,1536,0.0042275554604000514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,2048,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,2560,0.020755555894639757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,1024,0.0035111109415690103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,2048,0.018232888645595975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,1536,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,1536,0.015840889679061044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,1024,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,768,0.0031386667655573953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,6144,0.03738311264250014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,1024,0.013077333569526672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,768,0.007589333587222629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,512,0.00281866660548581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,768,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,768,0.012358222570684222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,3072,0.012905778156386482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,256,0.00250133333934678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,512,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,512,0.010783111055692038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,256,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,2048,0.010551999840471478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,256,0.010321777727868822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,1536,0.008363555702898238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,128,0.006945778098371293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,64,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,32,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,128,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,1024,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,65536,0.05200711223814222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,128,0.00814666681819492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,64,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,32,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,51200,0.041839109526740186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,65536,0.10336177878909641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,65536,0.17265866862403023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,512,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,51200,0.13566133711073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,51200,0.08341066704856025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,65536,0.31614934073554146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,16384,0.02607288956642151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,128,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,12288,0.012764444781674279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,16384,0.04620000057750278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,51200,0.24979376792907715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,10240,0.011390222443474663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,12288,0.021377777059872944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,16384,0.08692444695366754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,12288,0.035220444202423096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,8192,0.01201777739657296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,10240,0.030363556411531236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,10240,0.01965688831276364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,7168,0.011039999624093374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,12288,0.06666133138868544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,8192,0.01550044450494978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,10240,0.0573822193675571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,6144,0.009728000395827824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,8192,0.02494488822089301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,7168,0.014442667365074158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,8192,0.04715111189418369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,7168,0.022666666242811415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,5120,0.008382221890820397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,6144,0.01312622262371911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,16384,0.012050666742854647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,6144,0.02030933399995168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,4096,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,7168,0.04173688756095039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,5120,0.012070222033394707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,6144,0.03780622283617655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,5120,0.017656000124083627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,4096,0.010837333069907295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,4096,0.015232889188660515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,3072,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,5120,0.03198844525549147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,3584,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,4096,0.027412444353103638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,3584,0.014103111293580798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,3072,0.00998577806684706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,3072,0.012783110969596438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,2048,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,2560,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,2560,0.01148799972401725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,3072,0.023030221462249756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,1536,0.004214222232500712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,2048,0.008984888593355814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,2560,0.02091288897726271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,2048,0.010264889233642155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,1024,0.0032071111102898917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,1536,0.008360000120268928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,1536,0.008719111482302347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,2048,0.018238221605618794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,768,0.0028488888508743713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,1024,0.007527111305130853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,1024,0.008016000191370646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,3584,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,1536,0.015805333852767944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,512,0.00253866674999396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,768,0.007690666450394525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,768,0.006908444480763541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,256,0.0022906667242447534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,1024,0.01348355578051673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,2560,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,512,0.007015110717879401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,512,0.005970666805903117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,128,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,768,0.012457778056462606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,256,0.006913777854707506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,256,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,512,0.011368888947698804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,32,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,128,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,128,0.0052266667286554975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,65536,0.04555288950602213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,256,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,51200,0.03726666503482395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,16384,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,64,0.005197333378924264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,32,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,12288,0.012080888781282636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,10240,0.010603555374675326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,65536,0.17231199476453993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,51200,0.1353208886252509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,8192,0.009272889130645329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,16384,0.04591466652022468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,7168,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,12288,0.03479911221398248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,10240,0.02981511089536879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,6144,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,5120,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,8192,0.02479733361138238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,7168,0.02250755495495266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,4096,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,3584,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,6144,0.020120888948440552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,3072,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,64,0.0019617778145604664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,2560,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,4096,0.015136889285511441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,3584,0.014144000079896716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,3072,0.012732444538010491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,1536,0.003879111260175705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,128,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,1024,0.003469333259595765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,2560,0.011353777514563667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,768,0.0028862222615215513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,2048,0.010086221827401055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,512,0.002625777696569761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,1536,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,256,0.002293333411216736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,1024,0.007618666523032718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,128,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,768,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,64,0.001979555520746443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,512,0.005988444305128521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,32,0.001992000059949027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,256,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,65536,0.04673066735267639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,64,0.005057777795526716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,5120,0.01757244434621599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,16384,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,3584,0.02539288832081689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,12288,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,2048,0.004576888763242298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,65536,0.1716675493452284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,51200,0.13523111078474256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,16384,0.04573422339227465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,8192,0.011164444188276926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,7168,0.01015644437736935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,12288,0.03500444359249539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,10240,0.029743999242782593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,8192,0.024736000431908503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,5120,0.008579555484983657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,6144,0.020191111498408847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,4096,0.007585778004593319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,128,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,3584,0.007353777686754863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,51200,0.03570844398604499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,3072,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,4096,0.015167999598715039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,2560,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,3584,0.014092443717850579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,10240,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,2048,0.004351110921965705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,1536,0.003702222059170405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,2560,0.01142133358452055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,1024,0.0031662223239739737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,2048,0.010057777994208867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,6144,0.009454222189055549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,768,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,1536,0.008681777450773451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,512,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,256,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,1024,0.0075902218619982404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,7168,0.022442667020691767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,128,0.0019475554840432273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,64,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,768,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,512,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,32,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,256,0.005349333501524395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,5120,0.017681777477264404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,64,0.005099555684460534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,32,0.004972444640265571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,32,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,3072,0.012797333300113678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,128,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,16384,0.7916888660854764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,12288,1.9149786631266277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,16384,2.3346871270073786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,12288,0.5756790902879503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,16384,2.682143953111437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,10240,0.4922480053371853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,12288,1.857772403293186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,10240,1.4811058044433594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,8192,0.4005422327253554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,51200,2.877873738606771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,8192,1.1603582170274522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,10240,1.5255858103434246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,8192,1.2559697892930772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,7168,0.406513770421346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,6144,0.9577368630303277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,7168,0.9729546440972222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,6144,0.8761493364969889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,7168,1.056089825100369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,6144,0.3116675482855903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,5120,0.2662702136569553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,5120,0.8507395320468478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,4096,0.6106737984551324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,4096,0.21958933936225045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,5120,0.7578381962246366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,4096,0.5993608898586696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,3584,0.19592266612582734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,3584,0.6351991229587132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,3072,0.4379733403523763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,3072,0.1756604380077786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,3584,0.5377786424424913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,3072,0.4419457647535536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,2560,0.420288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,51200,7.38043212890625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,2560,0.15334667099846735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,2048,0.3434880044725206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,2048,0.14454132980770534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,1536,0.2884417904747857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,1024,0.15954577922821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,1536,0.104221330748664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,768,0.1388782262802124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,1024,0.09208266602622138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,1024,0.22860799895392525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,512,0.10181866751776801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,768,0.1448835531870524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,1536,0.24990754657321504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,256,0.07230933507283528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,512,0.07108977768156263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,512,0.12610399723052979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,128,0.05814222494761149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,256,0.1128951046201918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,128,0.0588888857099745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,64,0.051939556996027626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,2560,0.4374969005584717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,128,0.10525155729717678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,32,0.05388799972004361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,2048,0.3010088867611355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,64,0.10706222057342529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,32,0.10637778043746948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,768,0.07767555448744032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,256,0.06025155385335287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,51200,10.182909647623697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,16384,1.9273963504367406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,16384,2.053826649983724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,51200,2.346352047390408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,16384,0.6055875354342991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,65536,2.789680904812283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,12288,0.5053760210673014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,12288,1.4263155195448134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,10240,1.2205466164482963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,12288,1.577647950914171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,51200,5.679620530870225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,10240,0.38887998792860246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,10240,1.2759182188245985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,8192,1.112888018290202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,8192,0.3143528832329644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,7168,0.8277137544420031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,8192,0.953770637512207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,65536,7.397745768229167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,6144,0.2490631209479438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,7168,0.3036746713850233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,7168,0.8430328898959689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,6144,0.7249884605407715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,5120,0.2071253326204088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,4096,0.4482560157775879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,5120,0.6101075808207194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,6144,0.7121191024780273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,5120,0.587023999955919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,4096,0.17218044069078234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,51200,7.8230226304796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,4096,0.5367715623643663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,3584,0.42210666338602704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,3584,0.15557866626315647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,3072,0.34157776832580566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,3072,0.13719022274017334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,2560,0.27706223063998753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,3072,0.3527217706044515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,2560,0.11983111169603135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,2048,0.26528533299763996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,1536,0.17710933420393202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,2560,0.31767198774549693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,2048,0.10149066978030735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,3584,0.4040817684597439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,1024,0.12335377269321018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,1536,0.08226133055157132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,2048,0.2397893269856771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,1536,0.18951022624969482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,1024,0.06658222277959187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,512,0.078366219997406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,256,0.05353066656324598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,1024,0.14567822880215114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,512,0.054413331879509814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,768,0.11733955807156032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,128,0.045114666223526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,512,0.09899110926522149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,256,0.04931022061241997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,64,0.03930400146378411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,256,0.08819022443559434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,128,0.04763911167780558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,32,0.040375110175874494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,128,0.08468266990449692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,64,0.08381511105431451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,32,0.08454844686720107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,65536,10.453923543294271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,768,0.09975733359654744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,768,0.059212446212768555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,16384,0.5512586699591743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,65536,0.8777262369791666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,51200,0.6919066641065809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,16384,0.264540433883667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,12288,0.43262577056884766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,51200,2.1424187554253473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,12288,0.18134044276343453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,65536,2.4702186584472656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,12288,0.48958932028876406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,10240,0.36342400974697536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,10240,0.15709600183698866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,8192,0.29685600598653156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,8192,0.1070435576968723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,8192,0.3065440124935574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,51200,2.447376039293077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,7168,0.29223643408881295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,7168,0.09557422002156575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,6144,0.21050311459435356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,6144,0.09035199880599976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,7168,0.2779902352227105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,5120,0.19840178224775526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,65536,3.3390577104356556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,6144,0.23164799478318956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,5120,0.08123377958933513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,16384,0.6514506869845921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,4096,0.0610497792561849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,5120,0.1989964379204644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,3072,0.12169778347015381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,4096,0.16285067134433323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,3584,0.05475377705362108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,2560,0.09921866655349731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,3584,0.13816444079081217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,3072,0.048707554737726845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,3072,0.1184320052464803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,2048,0.08408621946970622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,2560,0.04405866728888618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,1536,0.06241244739956326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,2560,0.10052977667914496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,10240,0.42793599764506024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,1536,0.030206223328908283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,2048,0.08458133538564046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,768,0.030958222018347845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,1536,0.06875199741787381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,1024,0.024719999896155462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,1024,0.05008177624808418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,512,0.02533688810136583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,768,0.022340445054901972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,256,0.01791733337773217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,768,0.04081333345837063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,4096,0.15233066346910265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,512,0.020184889435768127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,128,0.01420177850458357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,3584,0.13109510474734837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,512,0.033392889632119074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,256,0.01864177816443973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,64,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,32,0.012485333614879184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,128,0.018579555882347953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,256,0.029979556798934937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,128,0.02881244487232632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,64,0.02847111225128174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,32,0.02883466747072008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,1024,0.04115022222201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,16384,0.5005840195549859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,65536,0.6944008933173286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,51200,0.548326227400038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,16384,0.18688178062438965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,12288,0.4072391192118327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,51200,1.673999998304579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,2048,0.038014223178227745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,16384,0.5193751123216417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,12288,0.12363199392954509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,10240,0.10547022024790446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,10240,0.3250417709350586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,12288,0.36996889114379883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,51200,1.9499458736843531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,8192,0.26922133233812123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,10240,0.28599733776516384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,7168,0.2013333373599582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,8192,0.09691288736131455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,8192,0.23131465911865234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,6144,0.21237599849700928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,7168,0.21855200661553276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,6144,0.06751466459698148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,5120,0.17488622665405273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,6144,0.20896266566382515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,5120,0.057168887721167676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,4096,0.1408782270219591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,5120,0.14608977900611028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,3584,0.11400977770487468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,4096,0.04773155517048306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,3072,0.0986586676703559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,3584,0.04255022274123298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,3584,0.10579111178716023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,3072,0.03812355465359158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,2560,0.08298399713304308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,3072,0.09243200222651164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,2048,0.05839288896984524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,2560,0.03374844458368089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,65536,2.3945689731174045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,2560,0.07974666357040405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,1536,0.05038933290375603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,7168,0.09507911072836982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,2048,0.06492177645365398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,1024,0.036502222220102944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,1536,0.02409511142306858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,1536,0.05180355575349596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,768,0.02576622201336755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,1024,0.020054222808943856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,512,0.020279111133681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,1024,0.03944089015324911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,768,0.018537777993414137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,65536,1.8646621704101562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,256,0.014351111319330005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,768,0.032313777340783015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,512,0.01686844395266639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,128,0.01070755554570092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,256,0.015408888459205627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,64,0.009692444569534725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,512,0.02664622167746226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,32,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,256,0.02352977792421977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,128,0.022668444448047217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,64,0.02270577847957611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,32,0.022678222921159532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,2048,0.029327998558680218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,4096,0.13378222783406576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,65536,0.5746320088704427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,128,0.015499555402331881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,51200,1.2166151470608182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,16384,0.13707821898990208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,16384,0.432985782623291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,65536,1.5901226467556422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,51200,0.45594220691257054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,16384,0.39240890079074436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,12288,0.314155552122328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,12288,0.12087466981675889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,10240,0.09143200185563828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,10240,0.27185156610276967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,51200,1.5884684456719294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,12288,0.286519103580051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,65536,2.0001903110080295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,8192,0.22156711419423422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,10240,0.26762133174472386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,7168,0.17638756169213188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,7168,0.06467199987835355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,6144,0.1591271162033081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,8192,0.2103466722700331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,5120,0.12690667311350504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,6144,0.056672890981038414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,7168,0.18596088886260986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,4096,0.11278666390313043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,5120,0.05007644494374593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,3584,0.09626044167412652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,6144,0.1616657839881049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,4096,0.04047200083732605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,5120,0.12335111035241021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,3072,0.0758320026927524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,4096,0.10309066375096639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,2560,0.07187733385297987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,3584,0.09400533305274116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,3072,0.03273155623012119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,2048,0.05383733246061537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,1536,0.039819555150138006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,2560,0.029599110285441082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,3072,0.07923911015192668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,2048,0.0251555558707979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,2560,0.067885332637363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,1536,0.021292444732454088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,2048,0.05580621957778931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,768,0.022175111704402502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,1024,0.01773422294192844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,1536,0.044533332188924156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,512,0.01591466698381636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,1024,0.034429334931903414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,768,0.01650044487582313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,8192,0.07357066869735718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,256,0.014988443917698331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,512,0.014904000692897372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,768,0.02815377712249756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,128,0.00963199966483646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,64,0.008554666406578487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,256,0.013879111243618859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,128,0.013795555465751223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,32,0.00886222223440806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,256,0.020319999919997323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,128,0.019659555620617337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,64,0.01965777741538154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,3584,0.03640444411171807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,32,0.01962577799956004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,1024,0.02793333265516493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,65536,0.48348357942369247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,16384,0.2873271041446262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,51200,0.3754986657036676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,65536,1.0751351250542533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,512,0.023011555274327595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,51200,0.9637022018432617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,16384,0.11123732725779216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,12288,0.24039554595947266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,16384,0.38762667444017196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,10240,0.19193866517808703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,12288,0.11040266354878743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,51200,1.1556417677137587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,12288,0.23378311263190377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,8192,0.13339466518825954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,10240,0.07397155629263984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,65536,1.6583733028835719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,7168,0.14075556066301134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,8192,0.05860177675882975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,10240,0.23449156019422743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,6144,0.11411644352806939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,7168,0.05334933267699348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,8192,0.18915643956926134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,5120,0.10060977935791016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,4096,0.08413244618309869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,6144,0.045688887437184654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,5120,0.03934577769703335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,3584,0.0714977780977885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,6144,0.12101511160532634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,4096,0.035368889570236206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,5120,0.10812711053424412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,3072,0.06170666880077786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,4096,0.08330577611923218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,2560,0.05076711045371162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,3072,0.02836444311671787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,3584,0.07369333505630493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,2048,0.039968000517951116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,2560,0.024687111377716064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,3072,0.06601778003904554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,1536,0.02986577815479702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,2560,0.055608888467152916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,2048,0.021737777524524268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,1024,0.022554667459593877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,2048,0.04628800021277534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,1536,0.018345778187115986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,768,0.01815555493036906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,1536,0.037909332248899676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,512,0.013452444639470843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,1024,0.015631111131774057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,768,0.014142221874660917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,1024,0.02904355525970459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,256,0.014085332552591959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,768,0.0239982224173016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,512,0.013232889274756113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,128,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,7168,0.14716711309221056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,512,0.019309333629078336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,256,0.012263110942310758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,64,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,32,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,128,0.012366222010718452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,256,0.017254221770498488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,128,0.01677688956260681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,3584,0.030057777961095173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,64,0.016743999388482835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,32,0.016619554824299283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,16384,0.26062220997280544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,65536,0.4235164324442546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,51200,0.33690844641791445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,16384,0.27877243359883624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,16384,0.10133955876032512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,12288,0.22475110159979927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,51200,0.8047546810574002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,10240,0.19440710544586182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,12288,0.07859910859002008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,65536,1.2222809261745877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,10240,0.08534844716389973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,12288,0.2168133391274346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,8192,0.1455048852496677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,51200,1.0125128428141277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,7168,0.13921777407328287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,65536,1.420871098836263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,6144,0.11777688397301568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,7168,0.059162669711642794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,8192,0.14079999923706055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,7168,0.1254311137729221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,5120,0.08381066719690959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,6144,0.04968888892067803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,4096,0.07703999678293864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,6144,0.11430844995710586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,3584,0.06786221928066678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,4096,0.03686044282383389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,5120,0.10172622071372138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,3072,0.05818577607472738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,4096,0.0770160026020474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,3584,0.02834933333926731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,2560,0.046032889021767505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,3584,0.06881422466701932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,3072,0.02641777859793769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,2048,0.036066667901145086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,3072,0.057893335819244385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,1536,0.027279999521043565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,2560,0.023853333459960088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,2560,0.05006666647063362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,2048,0.020346666375796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,1024,0.01974311139848497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,2048,0.04253066579500834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,1536,0.017274666163656447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,10240,0.22772532039218477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,8192,0.06564533048205905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,768,0.016559999850061204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,1536,0.03381422162055969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,1024,0.014702222413486905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,1024,0.02609066665172577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,768,0.013463999662134381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,256,0.008969777988062965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,512,0.012657778130637275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,768,0.021798221601380244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,512,0.01756533318095737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,5120,0.03632888860172696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,256,0.011758222348160215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,64,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,256,0.015795555379655626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,128,0.011542222566074796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,32,0.008276444342401292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,128,0.015497777197096082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,64,0.015429332852363586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,32,0.015415999624464246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,65536,0.3698408868577745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,512,0.012392000191741519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,128,0.007739555504586961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,65536,0.9884968863593208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,16384,0.2385164366828071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,51200,0.8411964310540093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,51200,0.2894471221499973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,12288,0.18845689296722412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,65536,1.106047100490994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,16384,0.0896257758140564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,51200,0.9491706424289279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,12288,0.07101333141326904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,10240,0.17639022403293184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,16384,0.2406675550672743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,8192,0.13480267259809706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,12288,0.2088151110543145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,10240,0.06033777528338962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,7168,0.117594665951199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,8192,0.05860800213283963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,6144,0.09909866915808783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,10240,0.1552977826860216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,7168,0.04320533408059014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,8192,0.12767822212643093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,5120,0.09260533253351848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,6144,0.03783466749721103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,4096,0.0662684440612793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,7168,0.13601333565182155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,6144,0.09731200006273057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,3584,0.05709244145287407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,4096,0.029440889755884807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,5120,0.08240889178382026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,3584,0.027797334724002417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,2560,0.043559110826916166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,4096,0.06547644403245714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,3072,0.02460977766248915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,2048,0.033013333876927696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,3072,0.05054666598637899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,2560,0.021278222401936848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,1536,0.025303110480308533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,2048,0.018664888209766813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,2560,0.044102221727371216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,1024,0.016579555140601266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,1536,0.01609422266483307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,2048,0.03907377852333917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,768,0.013360000318951078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,1024,0.013736888766288757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,512,0.010793777803579966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,1024,0.02328799996111128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,768,0.013007111019558378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,256,0.00796622203456031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,768,0.01953866746690538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,512,0.011751111182901593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,128,0.007251555720965068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,512,0.016072000066439312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,256,0.011111111276679568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,64,0.006212444355090459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,256,0.014478221535682678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,5120,0.03297244509061178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,128,0.011071110765139261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,32,0.006364444477690591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,3072,0.04751200146145291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,128,0.013759999639458127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,64,0.013783999615245394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,3584,0.05783466498057047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,32,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,65536,0.3220213254292806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,16384,0.1852568917804294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,51200,0.25085422727796763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,1536,0.030305776331159804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,65536,0.7564871046278211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,51200,0.6712142096625434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,12288,0.1583013269636366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,16384,0.07747999827067058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,10240,0.1481164428922865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,16384,0.21171377764807808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,12288,0.06531644529766507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,51200,0.736871984269884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,65536,0.9847662183973523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,10240,0.05215644505288866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,8192,0.11745599905649821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,7168,0.09112622340520223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,10240,0.13767911328209773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,8192,0.044028444422615894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,7168,0.04777600036727058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,8192,0.12083378103044297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,5120,0.07286044624116686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,6144,0.03522400061289469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,7168,0.09384355280134414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,4096,0.05475911166932848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,6144,0.08538577953974406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,5120,0.03747289048300849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,3584,0.0451564457681444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,4096,0.02551377813021342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,5120,0.06945333215925428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,3072,0.041519112057156034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,4096,0.05611644188563029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,3584,0.023165333602163527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,2560,0.02997866604063246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,3584,0.06184266673194038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,3072,0.021290666527218286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,2048,0.026820444398456152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,3072,0.04487733377350701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,2560,0.018819555640220642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,1536,0.02017688916789161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,2560,0.038930667771233454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,2048,0.01698844465944502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,1024,0.01496355566713545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,12288,0.15536977185143366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,1536,0.0144533332851198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,2048,0.0329075555006663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,768,0.012039999994966718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,512,0.009139555195967356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,1024,0.012719111310111152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,1536,0.02695200012789832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,768,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,6144,0.07875911394755046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,256,0.007755555212497711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,1024,0.02051733268631829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,128,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,512,0.011057777537239922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,768,0.017278222574128043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,64,0.005479111025730769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,512,0.014151111245155334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,256,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,32,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,128,0.010437332921557956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,256,0.012892444100644855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,128,0.012423111332787408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,64,0.012431999875439538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,32,0.012390221986505719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,16384,0.14937244521247015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,65536,0.28298576672871906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,51200,0.46884531444973415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,51200,0.22403645515441895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,65536,0.5975786844889323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,12288,0.1178800000084771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,16384,0.06809244553248088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,12288,0.06327466832266913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,16384,0.17069866922166613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,10240,0.11270756191677517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,51200,0.6069404284159342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,65536,0.7761991288926867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,8192,0.069069332546658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,7168,0.06906311379538642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,10240,0.05570755402247111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,6144,0.05690222316318088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,8192,0.0443102220694224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,10240,0.10977066887749566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,5120,0.054267555475234985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,8192,0.08847644593980576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,6144,0.030917333232031927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,7168,0.0800186660554674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,4096,0.03785689009560479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,3584,0.032481776343451604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,5120,0.028014222780863445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,4096,0.02219644520017836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,5120,0.05796000030305651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,3072,0.03056710958480835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,3584,0.021168000168270532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,4096,0.049750222100151904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,2560,0.0262435558769438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,3584,0.04294044441647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,3072,0.018279999494552612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,2048,0.020637333393096924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,2560,0.017204445269372728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,1536,0.016919111212094624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,2560,0.034019556310441755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,2048,0.014710222681363424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,1024,0.012786666552225748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,2048,0.027948445743984644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,1536,0.013243555194801755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,768,0.010861333045694562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,12288,0.13124799728393555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,1024,0.011382222175598145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,512,0.008440000315507254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,768,0.01073333372672399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,7168,0.03293955657217238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,768,0.014875556031862894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,256,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,6144,0.07633333073721991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,512,0.012454222473833295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,128,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,256,0.009639110830095079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,256,0.011370666325092316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,64,0.00453599997692638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,128,0.009739555418491364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,32,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,128,0.011061333119869232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,3072,0.03807555635770162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,64,0.010739555789364709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,32,0.011032000184059143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,1536,0.02312444481584761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,1024,0.017626666360431247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,65536,0.5329359902275933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,65536,0.2590968873765734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,16384,0.1488400035434299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,51200,0.42972710397508407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,512,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,51200,0.20672267013125947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,12288,0.10901600122451782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,65536,0.6829279793633355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,16384,0.07595733139250013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,10240,0.09811555676990086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,12288,0.049096888966030545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,51200,0.5328319867451986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,10240,0.043338666359583534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,12288,0.11690578195783828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,8192,0.034373333056767784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,10240,0.0995146698421902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,6144,0.055270221498277455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,8192,0.0992586678928799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,5120,0.04692799846331278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,7168,0.07127288977305095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,6144,0.027871112028757732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,4096,0.035402665535608925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,6144,0.061850666999816895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,5120,0.054903109868367515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,3584,0.03204355637232462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,4096,0.021036444438828364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,4096,0.043040888177024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,3072,0.026722666290071275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,3584,0.019295111298561096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,16384,0.16575022538503012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,2560,0.02312711046801673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,8192,0.0752568907207913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,3072,0.017637333936161466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,3072,0.034536888202031456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,2048,0.01853511068556044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,7168,0.06388711267047457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,1536,0.015268445014953613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,2560,0.016200888488027785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,2560,0.02983377708329095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,7168,0.03129244512981839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,1024,0.01106844428512785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,2048,0.025255999631351892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,1536,0.012798222402731577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,768,0.009339555270142024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,1536,0.0206657780541314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,1024,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,512,0.007145778172545963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,768,0.010681777364677854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,1024,0.01626133256488376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,512,0.009682666924264696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,512,0.011587555209795633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,256,0.009353777600659264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,128,0.005108444641033809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,256,0.010528888967302112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,64,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,128,0.00925422211488088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,3584,0.03891466723548041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,32,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,128,0.01001422190003925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,64,0.010061333576838175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,32,0.010082667072614035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,2048,0.01440711153878106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,65536,0.2164320018556383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,65536,0.539830207824707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,5120,0.024736889534526404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,16384,0.11943377388848199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,768,0.013926222920417786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,65536,0.5964978006150987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,12288,0.09082133240169948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,51200,0.1726844443215264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,16384,0.06712710857391357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,10240,0.08278933498594496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,51200,0.46366135279337567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,16384,0.13896000385284424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,12288,0.04281333420011732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,12288,0.10953511132134332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,10240,0.03922933340072632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,7168,0.05550666650136312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,8192,0.03103999959097968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,10240,0.09677422046661377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,6144,0.04836888776885139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,8192,0.07265155845218234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,7168,0.0296382208665212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,5120,0.036702222294277616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,6144,0.025029333101378545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,7168,0.06576444705327351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,6144,0.056045333544413246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,4096,0.0318195554945204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,5120,0.0217848883734809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,3584,0.02807733416557312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,4096,0.01928622192806668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,256,0.005547555370463266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,51200,0.4008479913075765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,4096,0.03999200132158067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,3072,0.024280889166726008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,2560,0.020799110333124798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,3584,0.01721600029203627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,3072,0.015455111861228943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,3584,0.03604177633921305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,2048,0.016470222009552848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,2560,0.014448889427714877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,3072,0.031552000178231135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,1536,0.013247110777431063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,2560,0.02774488925933838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,2048,0.01313511116637124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,1024,0.009662222531106737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,1536,0.011708444191349877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,2048,0.02308977809217241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,768,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,1024,0.01033688916100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,512,0.006144000010357962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,1024,0.014915555715560913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,768,0.009666666388511658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,256,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,768,0.013016888664828407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,512,0.009023110899660323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,128,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,512,0.011044444309340583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,256,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,64,0.004226666771703296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,256,0.009659555223253038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,128,0.008642666869693333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,32,0.004242666479614046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,128,0.009300444689061906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,64,0.009383111364311641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,32,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,65536,0.42166490024990505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,8192,0.06542488601472643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,51200,0.33029156261020237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,16384,0.09169244766235352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,1536,0.019640889432695176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,65536,0.5069875717163086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,51200,0.1592693328857422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,12288,0.08196622133255005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,51200,0.3993226687113444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,16384,0.051254222790400185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,10240,0.06756622261471219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,12288,0.04389866524272495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,8192,0.0543084442615509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,10240,0.03568177753024631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,12288,0.09102221992280747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,5120,0.04786844385994805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,10240,0.07748799853854708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,6144,0.03786399960517883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,8192,0.06531733274459839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,7168,0.02659999993112352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,5120,0.03465155429310269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,7168,0.05562488900290596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,6144,0.0244835548930698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,4096,0.02732266651259528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,6144,0.04888800117704603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,5120,0.020356444848908316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,65536,0.20076799392700195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,3584,0.022282666630215112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,5120,0.04190488987498813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,3072,0.02139733400609758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,4096,0.01796533332930671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,4096,0.034620443979899086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,2560,0.01789866718981001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,3072,0.015150222513410779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,3072,0.027777777777777776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,16384,0.12957866986592612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,2048,0.014743111199802823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,2560,0.02418311105834113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,1536,0.012074666718641916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,2048,0.012705778082211813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,7168,0.0464915566974216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,8192,0.0284897784392039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,1024,0.009072889056470659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,2048,0.02032177812523312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,768,0.007818666597207388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,1536,0.011217777927716574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,1024,0.009941333697901832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,1536,0.016872889465755887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,512,0.0063884444534778595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,768,0.009444444543785518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,1024,0.013474666410022311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,256,0.005096000101831225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,768,0.011759111450778114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,128,0.004254222330119875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,256,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,512,0.009842666486899057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,256,0.009179555707507664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,128,0.008360000120268928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,64,0.0038720000949170855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,3584,0.016442666451136272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,128,0.00870666652917862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,32,0.0042151109211974675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,3584,0.031217777066760596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,2560,0.013838222457302941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,64,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,32,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,65536,0.3252844545576307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,65536,0.15723821851942274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,16384,0.08137866523530748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,51200,0.20870844523111978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,51200,0.12389955255720352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,12288,0.05577244361241659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,65536,0.43770578172471786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,16384,0.04719466633266873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,10240,0.05075733198059929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,12288,0.03336888882848952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,51200,0.34764801131354434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,8192,0.03864266806178623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,10240,0.02903555499182807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,12288,0.0799600018395318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,7168,0.03368799885114034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,10240,0.07631644275453356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,6144,0.030951112508773804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,8192,0.0557840002907647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,7168,0.021130666136741638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,5120,0.025397333833906386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,7168,0.049765331877602466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,6144,0.019366222951147292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,4096,0.020558221472634208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,6144,0.044608000251981944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,5120,0.01700444519519806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,3584,0.017110221915774874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,4096,0.01516888870133294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,4096,0.031238221459918555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,3072,0.01575377749072181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,3584,0.013805333111021252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,3584,0.02827822168668111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,16384,0.10531377792358398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,3072,0.01277511070171992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,3072,0.025031111306614343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,512,0.008903111020723978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,2048,0.011696000066068439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,1536,0.009372444616423713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,2560,0.021651556094487507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,8192,0.023043556345833674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,2048,0.01054133309258355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,1024,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,2048,0.018479110466109384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,768,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,1536,0.015388444066047668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,1024,0.009034666750166152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,512,0.0053102220926019884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,1024,0.012404444317022959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,768,0.008812444077597724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,768,0.010692444112565784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,256,0.0041866666740841335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,5120,0.03733866744571262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,512,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,256,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,256,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,64,0.0032053333189752367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,128,0.007721777591440413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,32,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,128,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,2560,0.014129777749379476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,2560,0.011763555308183035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,64,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,65536,0.24144355456034342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,32,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,51200,0.19539999961853027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,1536,0.009719111025333405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,65536,0.14705955982208252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,16384,0.06102044714821709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,51200,0.11724266741010879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,65536,0.3399617671966553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,12288,0.048804445399178394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,10240,0.03727466530270047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,16384,0.038859556118647255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,12288,0.03165866600142585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,51200,0.2678968906402588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,8192,0.03264977865748935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,128,0.0035546666218174826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,12288,0.06492888927459717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,7168,0.026114667455355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,512,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,10240,0.06120977799097697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,6144,0.022936888866954382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,8192,0.023108444280094568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,5120,0.01984177695380317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,7168,0.020527111159430612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,8192,0.04549244377348158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,6144,0.01827022267712487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,4096,0.016184889607959323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,7168,0.04059111078580221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,3584,0.014872888724009195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,5120,0.0162764439980189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,6144,0.036152889331181846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,4096,0.014471999473041959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,5120,0.030849777989917334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,3072,0.01311288856797748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,4096,0.025468443830808003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,2560,0.01142133358452055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,3584,0.023017777336968318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,3072,0.012611555556456247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,2048,0.009565333525339762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,3072,0.020643555455737643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,2560,0.011588444312413534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,1536,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,2560,0.018072888255119324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,2048,0.0103493332862854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,1024,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,2048,0.015508444772826301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,1536,0.009719111025333405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,768,0.005766222046481238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,1536,0.013069333301650153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,1024,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,1024,0.010810666614108615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,512,0.004856889032655292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,768,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,256,0.003913777983850903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,512,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,512,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,128,0.0032906666811969546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,256,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,3584,0.01370488852262497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,256,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,64,0.003165333428316646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,32,0.0032017777363459268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,128,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,128,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,64,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,65536,0.1917733351389567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,32,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,16384,0.08745421965916951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,51200,0.14931644333733454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,65536,0.1008488867017958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,10240,0.028169777658250596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,16384,0.054663111766179405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,65536,0.2844924396938748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,12288,0.03933066791958279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,51200,0.08132177591323853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,768,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,16384,0.027837332752015855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,10240,0.03296533226966858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,51200,0.22416265805562338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,16384,0.07320266962051392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,12288,0.022835556003782485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,10240,0.02085777786042955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,7168,0.02386222283045451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,12288,0.0561386677953932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,8192,0.01620444489849938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,10240,0.04715999960899353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,6144,0.020605333977275424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,8192,0.03906666570239597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,7168,0.014833778142929077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,4096,0.015325332681337992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,7168,0.03509155578083462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,6144,0.01387466655837165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,5120,0.012421333127551608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,3584,0.013139555851618448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,6144,0.030969778696695965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,4096,0.011128000087208219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,5120,0.026371555195914373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,3072,0.011988444460762871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,2560,0.010457778142558204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,3584,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,4096,0.022167111436525982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,3072,0.010229333407349056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,3584,0.020227554771635268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,2560,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,3072,0.01796000036928389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,1536,0.007159111400445302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,2560,0.016031111280123394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,2048,0.00925244473748737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,1024,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,2048,0.013999111122555204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,1536,0.008656889200210571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,768,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,1024,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,1536,0.011737777955002256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,512,0.00452533322903845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,8192,0.0270480000310474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,1024,0.009752000371615091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,256,0.0035048888789282907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,768,0.007618666523032718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,128,0.003170666595300039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,512,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,768,0.008413333031866286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,256,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,512,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,64,0.0028773333049482773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,5120,0.01756088932355245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,32,0.003195555466744635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,256,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,128,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,65536,0.1488764418496026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,64,0.0063822223908371395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,32,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,51200,0.10719733105765449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,16384,0.038823998636669584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,2048,0.008785777621799046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,65536,0.23936443858676484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,51200,0.0802346666653951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,12288,0.027152889304690894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,51200,0.18932710753546822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,16384,0.026966222458415564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,10240,0.02244088881545597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,12288,0.02295111119747162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,8192,0.01872711049185859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,12288,0.04805244339836968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,10240,0.020572443803151447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,7168,0.01703555550840166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,10240,0.04118577639261881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,8192,0.016197333733240765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,6144,0.015156444576051501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,8192,0.03429333368937174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,7168,0.014927999840842353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,128,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,7168,0.030633777379989624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,6144,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,4096,0.011384000380833944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,6144,0.026605332891146343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,5120,0.012392889294359418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,3584,0.009896000226338705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,65536,0.09962666696972317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,5120,0.023027555810080633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,4096,0.011147555377748279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,3072,0.008836444053384993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,4096,0.01961688862906562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,3584,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,2560,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,3584,0.01791466606987847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,3072,0.010082667072614035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,2048,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,1536,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,2560,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,16384,0.06252444452709623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,1024,0.004254222330119875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,2560,0.01443822185198466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,2048,0.009302222066455418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,768,0.00388711111413108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,1536,0.008674666285514832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,2048,0.012425777812798819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,1536,0.010632000035709806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,1024,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,768,0.007675555845101674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,1024,0.009047110875447592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,256,0.0028808888875775864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,512,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,768,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,128,0.0025724443710512584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,5120,0.013270222478442721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,512,0.0069120000633928515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,256,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,64,0.0027919999427265594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,32,0.0028880000528362063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,128,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,256,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,128,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,65536,0.11301155885060628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,64,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,51200,0.08703555663426717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,32,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,16384,0.033746666378445096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,65536,0.09885155492358738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,12288,0.02680622206793891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,51200,0.07981955342822604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,65536,0.20153421825832793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,16384,0.02625244524743822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,10240,0.022671111755900916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,3072,0.01616533266173469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,51200,0.159804450141059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,12288,0.02144711050722334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,8192,0.01902666687965393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,7168,0.017157332764731515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,10240,0.01959466603067186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,12288,0.04117777612474229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,6144,0.015232000086042615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,8192,0.015396444333924187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,10240,0.03525066706869338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,8192,0.029492444462246362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,5120,0.012676444318559436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,6144,0.013140444126394061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,7168,0.02604888876279195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,4096,0.010850666297806634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,5120,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,6144,0.023228444986873206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,3584,0.009733333355850643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,5120,0.020291555258962844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,4096,0.010784889260927836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,3072,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,3584,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,2560,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,3584,0.015558222929636637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,3072,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,2048,0.00664533343580034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,3072,0.014136888914638095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,2560,0.009319999979601966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,2560,0.012647111382749347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,1536,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,2048,0.008928889201747047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,16384,0.05299733413590325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,2048,0.011005333728260465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,1024,0.004522666749027041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,1536,0.008309333688682979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,512,0.0033804445217053094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,768,0.0041866666740841335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,1536,0.009587555295891231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,1024,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,512,0.003532444437344869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,7168,0.014255111416180929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,1024,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,256,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,768,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,128,0.0028213332924577924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,768,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,256,0.0069146665434042616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,512,0.0063475556671619415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,64,0.0024746666765875286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,256,0.00566133318675889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,32,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,128,0.006666666517655055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,4096,0.017121778594122995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,65536,0.06638755400975545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,51200,0.05379733443260193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,64,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,32,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,65536,0.0983591079711914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,16384,0.018533334136009216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,65536,0.1570604509777493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,12288,0.01551822159025404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,51200,0.07957866456773546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,16384,0.024806222981876794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,51200,0.12385066350301106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,10240,0.01694222291310628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,16384,0.042146666182412036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,12288,0.021078222327762183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,8192,0.01278133359220293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,10240,0.01960177719593048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,7168,0.011547555526097616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,10240,0.028171555863486394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,8192,0.015477332803938123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,512,0.007672000262472365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,8192,0.023414222730530634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,7168,0.014158222410413953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,5120,0.008731555607583787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,7168,0.021254221598307293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,6144,0.01309511148267322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,128,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,6144,0.019148444135983784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,4096,0.0075235557225015424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,5120,0.011797332929240333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,3584,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,5120,0.01682222220632765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,3072,0.006167999986145232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,4096,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,3584,0.01034311122364468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,2560,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,4096,0.014521777629852295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,3584,0.013432888521088494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,3072,0.009703111317422655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,2048,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,1536,0.004243555582231945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,2560,0.009280888570679558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,2048,0.008784888519181145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,1024,0.0035208890007601846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,2560,0.011112000379297467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,2048,0.009654222263230218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,1536,0.008292444050312042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,12288,0.032949331733915545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,768,0.0032044444233179092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,512,0.0028115556471877625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,1536,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,1024,0.007819555699825287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,256,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,768,0.007368889119890001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,6144,0.010044444766309526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,128,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,512,0.0070053330726093715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,768,0.006607111129495833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,64,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,256,0.006690666493442323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,32,0.0021991110924217436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,128,0.006830222076839871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,256,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,65536,0.04744088980886671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,128,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,64,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,51200,0.03823644585079617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,32,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,16384,0.01500266624821557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,65536,0.09833511379030015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,51200,0.07904444138209026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,65536,0.15661422411600748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,12288,0.012435555458068848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,3072,0.01221422188811832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,51200,0.12292888429429795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,16384,0.041504889726638794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,12288,0.02104088829623328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,12288,0.0325164430671268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,8192,0.008347555167145198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,10240,0.01957066688272688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,10240,0.027615999182065327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,1024,0.007358222372002072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,7168,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,8192,0.015339555011855232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,512,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,6144,0.00869955536392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,7168,0.014150222142537435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,8192,0.023231110639042322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,5120,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,7168,0.021007110675175984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,6144,0.01309333327743742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,4096,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,6144,0.01870844430393643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,3584,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,5120,0.016578666037983365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,4096,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,3584,0.010110222631030613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,3072,0.005623111294375525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,16384,0.0244186669588089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,3584,0.013192888763215808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,10240,0.010738666686746808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,3072,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,2048,0.004536889079544279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,2560,0.009312888814343346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,3072,0.012081777883900536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,2560,0.010788444843557147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,1536,0.0038373333712418876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,2048,0.008640888664457533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,1024,0.0031715554909573663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,2048,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,1536,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,768,0.0028693332440323303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,1536,0.008361777497662438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,1024,0.007721777591440413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,512,0.00254133323000537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,1024,0.00757155567407608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,768,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,5120,0.011749332977665795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,256,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,768,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,128,0.002161777681774563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,4096,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,512,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,64,0.0021537778278191886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,256,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,512,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,128,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,2560,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,65536,0.04407111141416761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,128,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,64,0.004990222139490975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,51200,0.033864001433054604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,16384,0.01238044434123569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,32,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,12288,0.01108266661564509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,10240,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,65536,0.15708800156911215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,16384,0.04121511181195577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,51200,0.12238400512271458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,8192,0.009312888814343346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,12288,0.03194488750563727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,7168,0.008608000146018134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,6144,0.00795022232664956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,10240,0.027425777581002977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,5120,0.00795911086930169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,4096,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,3584,0.00620177760720253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,8192,0.023039999935362074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,7168,0.020938666330443487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,3072,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,6144,0.018783110711309645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,2560,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,2048,0.004555555681387584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,4096,0.014440000057220459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,3584,0.013080000049538083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,32,0.0021439999755885866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,3072,0.012088889049159156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,256,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,1024,0.0032177778581778207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,2560,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,768,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,2048,0.009569777382744683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,512,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,256,0.002346666736735238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,128,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,1536,0.008618666893906063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,64,0.0018826667219400406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,768,0.006518222391605377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,32,0.001971555459830496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,256,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,128,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,65536,0.03940800163480971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,64,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,51200,0.03173955612712436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,32,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,16384,0.009383111364311641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,65536,0.15557066599527994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,12288,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,51200,0.12257688575320774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,5120,0.016536000702116225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,10240,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,16384,0.041133334239323936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,8192,0.006908444480763541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,7168,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,1536,0.0038515557017591265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,6144,0.00943911075592041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,12288,0.031981332434548274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,5120,0.009564444422721863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,10240,0.02747022277779049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,4096,0.008031999899281396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,7168,0.02093066606256697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,3584,0.007526222202512953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,1024,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,3072,0.006517333288987477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,512,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,5120,0.016528889536857605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,2560,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,2048,0.005127999931573868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,4096,0.014311111635631986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,1536,0.003798222376240624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,3584,0.013112000293201871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,1024,0.003137777869900068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,2560,0.010919111470381418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,2048,0.009722666607962714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,1536,0.008360000120268928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,512,0.002499555548032125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,1024,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,256,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,768,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,128,0.0019911111642917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,512,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,64,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,256,0.005315555466545953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,32,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,8192,0.023035556077957153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,128,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,6144,0.018886221779717337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,64,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,32,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,3072,0.012111110819710625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,768,0.002855111120475663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,16384,0.6933182080586752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,16384,2.238504833645291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,16384,2.5421600341796875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,12288,1.8554070790608723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,51200,2.6429004669189453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,16384,2.141265869140625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,12288,1.7875440385606554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,12288,0.532060464223226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,10240,0.47352001402113175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,10240,1.6019360224405925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,12288,1.590971522861057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,10240,1.4346560372246637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,8192,0.37608178456624347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,51200,7.130893283420139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,10240,1.3284062279595268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,8192,1.2029493119981554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,7168,0.3280124399397108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,7168,1.0643271340264213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,8192,0.9999075995551215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,8192,1.449634657965766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,7168,1.008286264207628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,6144,0.9476186964246961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,6144,0.28881245189242893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,7168,0.90102842119005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,6144,0.9045990837944878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,5120,0.7019724316067166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,5120,0.2462488810221354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,4096,0.5015920003255209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,5120,0.7568497657775879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,6144,0.7645831108093262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,5120,0.6289715766906738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,4096,0.6017644670274522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,4096,0.20393066936069063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,3584,0.5108364423116049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,3584,0.18375910653008354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,3584,0.5272675620185004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,3072,0.1610915528403388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,3584,0.4496311081780328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,3072,0.492752022213406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,51200,9.849801805284288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,2560,0.3518773449791803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,2560,0.1419795619116889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,2560,0.37817245059543186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,2048,0.2977004316118029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,4096,0.5252213478088379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,2560,0.33365776803758407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,3072,0.4133146603902181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,2048,0.2932079897986518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,1536,0.22075111336178252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,2048,0.1299253304799398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,1024,0.1508844428592258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,2048,0.2768062220679389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,1536,0.24662399291992188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,51200,7.1143603854709205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,1024,0.21795199977027047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,1024,0.0765520003106859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,1536,0.21996355056762695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,3072,0.40030399958292645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,512,0.09369422329796685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,1024,0.16739111476474336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,768,0.14318577448527017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,768,0.06760622395409478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,256,0.06276800235112508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,768,0.13495288954840765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,512,0.12191732724507649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,256,0.055380443731943764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,512,0.11219822035895453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,128,0.050015111764272056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,256,0.10830844773186578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,1536,0.09584533505969578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,256,0.0876400007141961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,64,0.046022223101721875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,128,0.053007112609015576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,32,0.04602311054865519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,128,0.10178933540980022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,128,0.06701333655251397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,64,0.10211555825339423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,32,0.10315466589397854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,512,0.061011552810668945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,768,0.12568977144029406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,65536,2.532057868109809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,51200,1.8878809611002605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,51200,5.384534200032552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,16384,1.9901608361138239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,16384,2.0704854329427085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,65536,7.017396291097005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,16384,0.6184115409851074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,12288,1.342593828837077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,51200,7.714644538031684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,12288,0.4275742106967502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,12288,1.3967919879489477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,16384,1.6415120230780706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,10240,1.145783954196506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,12288,1.2138053046332467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,10240,0.356478214263916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,65536,10.162346733940973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,10240,1.1960923936631944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,8192,0.908488909403483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,8192,0.2927475505405002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,51200,5.499170515272353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,10240,0.9590720070732964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,8192,0.9384400049845377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,7168,0.7897529072231717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,7168,0.2561173439025879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,8192,0.7831901974148221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,6144,0.6424471007453071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,7168,0.8030044237772623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,6144,0.6760533120897082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,6144,0.22377332051595053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,7168,0.693545765346951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,5120,0.5424168904622396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,5120,0.1942666636572944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,4096,0.4261031150817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,5120,0.6492337650722927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,4096,0.16224088933732775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,65536,7.1637013753255205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,5120,0.5024853282504612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,4096,0.46439557605319554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,3584,0.39653423097398544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,6144,0.5935137536790636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,3584,0.16301600138346353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,3072,0.312903987036811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,4096,0.4302693208058675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,3584,0.35796088642544216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,3072,0.33304444948832196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,3072,0.3100186718834771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,2048,0.1962826649347941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,2560,0.3031795554690891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,2048,0.0942266649670071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,2560,0.26624266306559247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,2048,0.23276800579494902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,1536,0.16770488686031768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,3584,0.3959191110399034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,3072,0.12890044848124185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,1536,0.07627466652128431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,1536,0.18396000067392984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,1024,0.12146933873494466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,2048,0.22106756104363334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,768,0.0885511106914944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,2560,0.2968497806125217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,1024,0.061526219050089516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,512,0.06520889202753703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,1024,0.13943644364674887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,768,0.054606222444110446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,768,0.11146222220526801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,1536,0.17260977956983778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,1024,0.13086399767133924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,256,0.04306755463282267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,512,0.049476444721221924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,128,0.03621866636806064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,768,0.1106577780511644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,256,0.04521244433191088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,256,0.0858888891008165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,512,0.09016799926757812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,128,0.04303200046221415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,64,0.032580445210138954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,32,0.03417422374089559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,256,0.07090844710667928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,128,0.05447466505898369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,128,0.08142933580610487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,64,0.08090666929880778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,32,0.08183199829525419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,2560,0.11310311158498128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,512,0.09634488821029663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,65536,0.7875279850429959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,16384,0.5508009062872993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,51200,2.1121465894911027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,65536,2.510501437717014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,51200,0.6177973217434353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,16384,0.1854106717639499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,16384,0.6157031059265137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,51200,2.545369678073459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,12288,0.5725839932759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,65536,3.340806325276693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,12288,0.14498133129543728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,16384,0.5329093403286403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,12288,0.4496720102098253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,12288,0.39359021186828613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,10240,0.12375911076863606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,8192,0.31134221288892955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,10240,0.410699552959866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,51200,1.893640094333225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,10240,0.33286576800876194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,7168,0.21672799852159289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,8192,0.2921244568294949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,7168,0.08847733338673909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,8192,0.270576900906033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,6144,0.17143377992841932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,7168,0.2830728954739041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,10240,0.3840000099605984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,6144,0.07835821972952949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,5120,0.18477511405944824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,6144,0.22755911615159777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,5120,0.06794044706556532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,6144,0.20715733369191489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,5120,0.1848355531692505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,65536,2.430349349975586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,5120,0.17352444595760771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,4096,0.05703555875354343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,4096,0.15069333712259927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,3584,0.14258400599161783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,3072,0.12206844488779704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,4096,0.1431679990556505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,3584,0.051017777787314526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,3584,0.1302151150173611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,2560,0.08768355846405029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,3072,0.04574044545491537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,7168,0.23527023527357313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,3072,0.11386222309536403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,2048,0.06826755735609266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,2560,0.04046222236421373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,3072,0.11248711744944255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,4096,0.12091022067599827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,2560,0.09804622332255046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,1536,0.04687644375695122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,2048,0.03489955597453647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,2048,0.0810462236404419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,2560,0.09678400225109524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,1024,0.0336133340994517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,1536,0.02860266632503933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,1536,0.0647546648979187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,2048,0.08217600319120619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,768,0.02994577752219306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,1024,0.02342133389578925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,1024,0.04909422331386142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,1536,0.06853777832455106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,512,0.024653333756658766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,768,0.021160889003011916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,1024,0.05255555444293552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,768,0.03965511255794101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,256,0.020414221617910597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,3584,0.12876178158654109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,512,0.0192248887485928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,512,0.03238933285077413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,128,0.014630221658282809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,256,0.017265778448846605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,256,0.029211554262373183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,512,0.03811466693878174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,64,0.012409778104888068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,32,0.012739555703269111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,128,0.017717333303557504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,256,0.03007733490731981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,128,0.023511999183230933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,8192,0.10408622026443481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,32,0.02779200010829502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,768,0.04555199874771965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,65536,0.611096011267768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,128,0.028183112541834515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,64,0.02830577890078227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,65536,1.8753395080566406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,51200,1.6068435245090062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,51200,0.4784862200419108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,16384,0.5047858026292589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,16384,0.16806399822235107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,65536,2.560286204020182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,16384,0.47228442298041445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,51200,1.7009627024332683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,16384,0.4344675540924072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,12288,0.11146399709913467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,12288,0.3601502312554254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,10240,0.28513511021931964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,65536,1.8713590833875868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,12288,0.3136320114135742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,10240,0.11165600352817112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,10240,0.28021955490112305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,8192,0.23502845234341094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,51200,1.4286382463243272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,8192,0.22658578554789224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,7168,0.20502755377027723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,10240,0.2605866591135661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,8192,0.07768444220225017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,6144,0.14503910806443956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,12288,0.3827635447184245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,8192,0.2284853458404541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,7168,0.07806133561664157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,7168,0.218121780289544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,5120,0.14321066273583308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,6144,0.16933866341908774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,6144,0.06184266673194038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,4096,0.11865066157446967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,7168,0.18745866086747912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,5120,0.0537377761469947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,6144,0.16149244043562147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,5120,0.1413831048541599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,3584,0.08709066443973118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,4096,0.04394577940305074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,4096,0.11385866006215413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,5120,0.14521422651078966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,3584,0.04010933306482103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,3584,0.10190666384167141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,4096,0.11378222041659886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,3072,0.03591377867592706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,3072,0.08905688921610515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,3584,0.10203911198510064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,2048,0.05679733223385281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,2560,0.03170844581392076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,3072,0.09162311421500312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,2560,0.07686755392286512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,1536,0.03939022289382087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,2560,0.07839288976457384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,2048,0.06313688887490167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,1024,0.02756711178355747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,2048,0.06791822115580241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,1536,0.022681777675946552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,1536,0.05048799845907423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,768,0.02181422213713328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,1024,0.019494222270117868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,1024,0.038551112016042076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,1536,0.05528444382879469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,512,0.01750577820671929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,3072,0.08760533067915176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,768,0.03165955675972833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,2560,0.07426133420732287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,1024,0.04403733213742574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,768,0.017846221725145977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,256,0.01811288959450192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,512,0.016117332710160148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,128,0.015259555644459195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,768,0.03702577617433336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,256,0.014734221829308404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,256,0.022840888963805303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,512,0.03194933467441135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,64,0.014051554931534661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,128,0.015110222829712762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,32,0.014463111758232117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,128,0.022056000100241765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,2048,0.02756622268093957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,256,0.02438755498992072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,128,0.019738665885395475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,64,0.022086222966512043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,32,0.021984888447655573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,65536,0.519245359632704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,512,0.025855110751258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,51200,1.1193564732869465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,16384,0.30076267984178334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,65536,1.5814257727728949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,51200,0.40358577834235293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,51200,1.3901537789238825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,16384,0.13946578237745497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,12288,0.29095999399820965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,65536,2.013959036933051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,16384,0.5217111375596788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,16384,0.34119555685255265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,10240,0.22395377688937715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,12288,0.2961164315541585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,8192,0.16145066420237222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,12288,0.25955022705925834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,10240,0.09271466732025146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,65536,1.5625101725260417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,8192,0.06887911425696479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,10240,0.22113777531517875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,7168,0.169795552889506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,8192,0.19387733936309814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,7168,0.06933866606818305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,8192,0.17732355329725477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,7168,0.1695528825124105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,6144,0.14494489298926458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,7168,0.15650310781266955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,51200,1.1766835318671334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,12288,0.10571733448240493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,6144,0.05624710851245456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,6144,0.14361866315205893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,5120,0.10206933154000176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,4096,0.0905288921462165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,5120,0.045570668247010976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,10240,0.23739467726813424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,6144,0.1365671157836914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,3584,0.07018399900860257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,5120,0.11938577228122288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,4096,0.09712000025643243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,3072,0.05984977881113688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,3584,0.03302844365437826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,4096,0.09868799977832371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,3584,0.0867884423997667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,2560,0.05659911367628309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,3072,0.030518223841985066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,3584,0.08733599715762669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,3072,0.07586755355199178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,2048,0.04138400157292684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,2560,0.02739199995994568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,3072,0.07742310894860162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,2560,0.06581155459086101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,1536,0.03386133246951633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,2048,0.023723555935753718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,2560,0.06865600082609388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,1536,0.04379466507169935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,1024,0.02349866595533159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,1536,0.020271999968422782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,4096,0.03753066725201077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,1536,0.0480915539794498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,1024,0.01721333298418257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,768,0.019222221440739103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,1024,0.03368355499373542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,512,0.014459555347760519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,768,0.015437333120240105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,1024,0.03865155577659607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,768,0.027687999937269423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,256,0.010681777364677854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,512,0.014159111513031853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,512,0.022312000393867493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,768,0.0328693323665195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,128,0.01019022199842665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,256,0.012874667015340594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,512,0.027842667367723253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,2048,0.05409511261516147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,64,0.008915555146005418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,128,0.01924800044960446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,128,0.013440888788965015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,32,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,2048,0.05738044447369046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,128,0.01755288905567593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,5120,0.11725155512491862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,32,0.01904000010755327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,65536,0.43443910280863446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,256,0.019855999284320407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,51200,0.9134675131903754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,256,0.022152000003390845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,65536,1.412338680691189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,64,0.019280888968043856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,16384,0.3345404465993245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,65536,1.6204391055636937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,51200,0.34128533469306094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,16384,0.3214213318294949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,16384,0.10337333546744452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,12288,0.26309511396620006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,12288,0.07923022243711683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,10240,0.21778489483727348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,16384,0.3009493350982666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,51200,1.2017528745863173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,12288,0.2698062260945638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,65536,1.2976213031344943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,10240,0.06851910882525973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,8192,0.16986578040652803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,12288,0.22759821679857042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,10240,0.1973635620541043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,8192,0.05557066864437527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,7168,0.1420995526843601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,51200,0.9929653803507487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,8192,0.15784888797336155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,10240,0.2000675598780314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,6144,0.12337333626217312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,7168,0.04965510964393616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,7168,0.13555999596913657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,8192,0.16410844855838352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,5120,0.0999164448844062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,6144,0.04420266548792521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,6144,0.12623733944363064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,7168,0.1385706663131714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,4096,0.08558222320344712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,5120,0.03872355487611558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,3584,0.0657048887676663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,5120,0.09937333398395115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,6144,0.1223306655883789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,4096,0.08074488904741076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,5120,0.10312444633907741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,3584,0.029280887709723577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,3072,0.06333066357506646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,3584,0.0721235540178087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,4096,0.08598311079872979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,2560,0.05138311121198866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,3072,0.026518222358491685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,3072,0.06278844674428304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,3584,0.07830488681793213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,2560,0.023790222075250413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,2048,0.040185779333114624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,2560,0.054416000843048096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,3072,0.06952622201707628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,1536,0.03040177623430888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,2048,0.020439111524158053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,2560,0.06097244554095798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,2048,0.04515999886724684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,1024,0.0222053329149882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,1536,0.01754133403301239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,1536,0.0369137790468004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,2048,0.05071555574735006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,768,0.01953866746690538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,1024,0.014702222413486905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,1024,0.028304000695546467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,1536,0.04269688990381029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,4096,0.03242933419015672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,512,0.01660711069901784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,768,0.013613333304723104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,1024,0.033695111672083534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,768,0.023245332969559565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,256,0.012147555748621622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,512,0.012480888929631976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,128,0.010330666270520952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,512,0.01906488835811615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,256,0.012016889121797351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,256,0.0170462214284473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,64,0.009297777381208207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,32,0.009692444569534725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,128,0.01625066664483812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,256,0.02029600077205234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,64,0.01633777717749278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,128,0.01609244445959727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,32,0.016173332929611206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,65536,0.3901617791917589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,65536,0.9246390660603842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,51200,0.7085199885898166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,512,0.024189333120981853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,128,0.012096000214417776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,16384,0.2782791190677219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,65536,1.4223386976453993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,51200,0.3026631143358019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,16384,0.2660586569044325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,16384,0.09931111335754395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,12288,0.20979555447896323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,65536,1.0648799472384982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,51200,1.0708017349243164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,12288,0.08045511113272773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,16384,0.2637102339002821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,10240,0.1761795547273424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,768,0.029465778006447688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,10240,0.06144444147745768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,8192,0.14042311244540745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,12288,0.2038853300942315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,10240,0.20328356160057914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,51200,0.8347555796305338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,7168,0.1277502245373196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,8192,0.062280886703067355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,8192,0.13589244418674046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,10240,0.17595110999213326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,7168,0.04481777879926893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,6144,0.11105155282550389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,5120,0.08582044310039944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,7168,0.1250391138924493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,8192,0.13966755072275797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,6144,0.0391813317934672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,6144,0.10491110881169637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,4096,0.07633599970075819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,5120,0.034522665871514216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,7168,0.13953067196740046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,3584,0.056233776940239795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,6144,0.1095599995719062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,4096,0.029321776496039495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,12288,0.1996622218026055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,4096,0.07169866561889648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,3072,0.05307288964589437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,3584,0.026752000053723652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,4096,0.07669511106279162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,2560,0.04602844516436259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,3584,0.06571377648247613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,3072,0.024150222539901733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,3072,0.05682488944795397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,2048,0.03469333383772109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,3072,0.06195466385947334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,2560,0.021963554951879714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,2560,0.04896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,1536,0.026478222674793665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,2048,0.01902577777703603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,2048,0.04083822170893351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,2560,0.05375377668274773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,1536,0.016650666793187458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,1536,0.03335466649797227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,2048,0.04672444529003567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,1024,0.020392888122134738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,5120,0.09316533141665989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,768,0.01664355562792884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,1024,0.013969777358902825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,1536,0.039297779401143394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,1024,0.02537511123551263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,512,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,768,0.021320889393488567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,1024,0.030908445517222088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,3584,0.07102488809161715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,256,0.009644444617960189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,512,0.012081777883900536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,512,0.017217778497272067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,768,0.02700800034734938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,128,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,256,0.01108000013563368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,512,0.02288088864750332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,256,0.015519110692871941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,64,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,128,0.014832889040311178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,128,0.011412444214026133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,32,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,256,0.018559111489189994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,128,0.015626667274369132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,32,0.014858666393491956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,5120,0.09242310788896348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,768,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,65536,0.3513911035325792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,65536,0.9159830941094292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,51200,0.7117644415961372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,16384,0.24560621049669054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,65536,1.0416942172580295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,51200,0.27765777375963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,64,0.01495733360449473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,16384,0.08559022347132365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,51200,0.823945787217882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,12288,0.1954266627629598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,12288,0.06488800048828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,10240,0.15900799963209364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,51200,0.7157324684990777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,65536,0.9437226189507378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,12288,0.17700089348687065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,16384,0.2343724568684896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,8192,0.13899732960595026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,10240,0.05658044417699178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,12288,0.17856710486941865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,7168,0.11933422088623047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,8192,0.0505600008699629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,10240,0.15060888396369085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,8192,0.12008178234100342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,6144,0.08956622415118748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,10240,0.15248355600568983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,7168,0.04141777753829956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,8192,0.12287110752529568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,5120,0.0864400002691481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,6144,0.03604266709751553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,6144,0.0920684470070733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,7168,0.11047377851274277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,4096,0.060066666867997914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,5120,0.030751999881532457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,5120,0.0780657794740465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,16384,0.24833867284986708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,6144,0.0950017770131429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,4096,0.02650044361750285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,5120,0.08140444755554199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,4096,0.06437333424886067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,3072,0.04895022180345324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,3584,0.02374222212367588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,4096,0.06975911060969035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,3072,0.021624000536070928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,2560,0.04115110966894362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,3072,0.05075822273890177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,3584,0.061533331871032715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,2048,0.03011111087269253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,2560,0.0196426659822464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,3072,0.05485599901941088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,7168,0.10748444663153754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,2560,0.04404800136884054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,1536,0.022818666365411546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,2048,0.017264000243610807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,2560,0.049118220806121826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,2048,0.03643644518322415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,1024,0.01629955569903056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,1536,0.01475644442770216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,1536,0.029959999852710303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,2048,0.04243910974926419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,768,0.01348888874053955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,3584,0.05424000157250298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,1024,0.012739555703269111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,1536,0.0348053342766232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,1024,0.0225582222143809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,768,0.011883555187119378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,1024,0.027849776877297297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,3584,0.056645333766937256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,256,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,512,0.011077333655622272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,512,0.015528000063366361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,768,0.024226667152510747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,128,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,256,0.010363555616802638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,256,0.014087999860445658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,512,0.02033511135313246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,64,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,128,0.010583110981517367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,256,0.017595556047227647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,128,0.013458666702111563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,32,0.006014222072230444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,64,0.013457777599493662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,128,0.014818666709793938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,32,0.013454222016864352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,512,0.011058666639857821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,65536,0.635575983259413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,768,0.01919644408755832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,65536,0.27501598993937176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,51200,0.45996978547837997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,16384,0.17709778414832222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,51200,0.2181839942932129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,16384,0.21086843808492026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,16384,0.0684408876630995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,51200,0.7092222107781304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,12288,0.11847555637359619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,16384,0.2082657814025879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,65536,0.8051324420505099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,12288,0.05388266510433621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,12288,0.17053243849012586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,10240,0.10667555862002903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,8192,0.08965600199169582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,10240,0.046077334218555026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,12288,0.1546142233742608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,10240,0.14628177218967015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,7168,0.07261421945359972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,8192,0.036890665690104164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,10240,0.12790666686164007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,8192,0.10587644577026367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,6144,0.07012799713346693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,65536,0.9749697579277886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,7168,0.033004442850748696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,5120,0.06291555696063571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,6144,0.029782222376929388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,7168,0.09277155664232041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,6144,0.07999733421537611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,7168,0.09238488806618585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,4096,0.04492444462246365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,5120,0.026520000563727483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,6144,0.08346399996015762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,3584,0.03730488816897074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,51200,0.6044711007012261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,4096,0.022932445009549458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,4096,0.0551457769340939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,3072,0.035311109489864774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,3584,0.02079288827048408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,3584,0.049696889188554555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,3072,0.044016887744267784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,3584,0.05419466561741299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,2560,0.028633776638242934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,8192,0.10640711254543728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,3072,0.04800533254941305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,2048,0.024650666448805068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,2560,0.03859377900759379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,2560,0.0169831116994222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,2560,0.04210755560133192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,1536,0.018947554959191214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,2048,0.01513155632548862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,2048,0.03224444389343262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,5120,0.07124355766508314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,2048,0.03682933251063029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,1024,0.013929777675204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,1536,0.02630399995379978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,4096,0.05928444200091892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,768,0.011599111060301462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,1536,0.030735111898846094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,1024,0.01995022263791826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,3072,0.01907999979125129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,512,0.008637333081828224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,768,0.01700355609258016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,1024,0.024842666255103216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,768,0.010753778119881948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,256,0.006416000011894438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,512,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,768,0.021290666527218286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,512,0.013898667362001208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,128,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,256,0.01242844429281023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,256,0.00962222201956643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,512,0.019488889310095046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,5120,0.06740266746944852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,64,0.00543200017677413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,32,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,128,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,128,0.0123831108212471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,256,0.01646933290693495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,1536,0.013456000222100152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,64,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,128,0.013665777941544851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,32,0.012091555529170565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,1024,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,65536,0.2536746660868327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,65536,0.6607404285007054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,51200,0.5132080184088813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,51200,0.199217783080207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,16384,0.19322755601671007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,65536,0.7449982431199816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,51200,0.5826844639248318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,12288,0.12782755162980822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,16384,0.07178488704893324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,16384,0.17980533175998262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,65536,0.7701395352681478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,51200,0.5899804433186849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,12288,0.05715466870201958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,12288,0.12945866584777832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,8192,0.0808488859070672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,10240,0.04567377765973409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,12288,0.14890044265323216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,10240,0.12548800309499106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,8192,0.034544888469907976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,8192,0.0875377787484063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,7168,0.07963022258546618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,8192,0.1034159991476271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,7168,0.03141599893569946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,6144,0.06655200322469075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,7168,0.07762310902277629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,7168,0.09124533335367839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,6144,0.028347555134031508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,5120,0.05440444416469998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,10240,0.11186577214135064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,6144,0.06703378094567193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,16384,0.19342488712734648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,4096,0.045477333996031016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,5120,0.02479644450876448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,6144,0.080494225025177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,3584,0.03941155473391215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,4096,0.02126666737927331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,5120,0.05748089154561361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,4096,0.0473653342988756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,3072,0.03375822305679321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,10240,0.11078133185704549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,3584,0.019543111324310303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,5120,0.06874933507707384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,3584,0.04262400004598829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,2560,0.028873778051800195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,4096,0.057371556758880615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,3072,0.018087110585636564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,2048,0.02409155501259698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,3072,0.03792444533771939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,2560,0.016176000237464905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,3584,0.052575111389160156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,1536,0.018623110320832994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,3072,0.0468355549706353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,2048,0.014555555250909595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,2048,0.027809778849283855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,1024,0.014750222365061441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,2560,0.04074666566318936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,1536,0.012997333374288348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,1536,0.02260355485810174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,2048,0.03523644473817613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,1024,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,1024,0.01740533279048072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,1536,0.028648889727062646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,512,0.008624888956546783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,768,0.014856000741322836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,768,0.010408889088365767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,256,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,512,0.012287110918098025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,768,0.020598222812016804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,512,0.01791822248035007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,256,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,256,0.011035555766688453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,128,0.005856888989607493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,128,0.009558222360081142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,256,0.014941333068741692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,64,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,128,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,2560,0.03289866778585646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,32,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,128,0.01259022206068039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,768,0.011680000358157687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,64,0.01069155583779017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,32,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,65536,0.45754845937093097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,1024,0.023644444015291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,65536,0.2522835466596815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,51200,0.4248515499962701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,512,0.009733333355850643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,65536,0.6479457749260796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,16384,0.14145067003038195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,51200,0.19399378034803602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,51200,0.5036808649698893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,12288,0.0978435542848375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,16384,0.06316888994640775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,16384,0.16679378350575766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,65536,0.609082645840115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,10240,0.09143733315997654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,12288,0.05208711160553826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,16384,0.15973599751790366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,12288,0.11395377582973903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,8192,0.0724053316646152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,10240,0.04260177744759453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,10240,0.09974310795466106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,12288,0.12169510788387722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,8192,0.034056001239352755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,7168,0.06230311261283027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,8192,0.0786328911781311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,10240,0.10991289218266805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,6144,0.051113777690463595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,7168,0.029896001021067303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,8192,0.08591644631491767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,7168,0.06875822279188368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,5120,0.03904622130923801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,6144,0.0270480000310474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,6144,0.061390221118927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,7168,0.07767555448744032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,4096,0.03459377752410041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,5120,0.02346755564212799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,5120,0.051334222157796226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,6144,0.06809155808554755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,3584,0.03048711021741231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,4096,0.020125332805845473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,4096,0.042862223254309766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,5120,0.0585048860973782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,3072,0.026408000124825373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,3584,0.018404444058736164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,3584,0.03856177793608771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,2560,0.021811554829279583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,51200,0.4826329019334581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,3072,0.017009778155220878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,3072,0.034330666065216064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,3584,0.04439289040035672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,2048,0.016943999462657504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,2560,0.014886221951908536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,2560,0.029575109481811523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,3072,0.0404151115152571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,1536,0.013474666410022311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,2048,0.013820444544156393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,2048,0.02498133314980401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,2560,0.03526044554180569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,1536,0.011747555600272285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,1536,0.020548444655206468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,2048,0.030996445152494643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,768,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,1024,0.01037333326207267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,1536,0.025565332836574976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,1024,0.015986666083335876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,512,0.007244444555706448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,768,0.013774221969975365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,768,0.009759999811649323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,1024,0.020282665888468426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,256,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,512,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,4096,0.0498595568868849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,768,0.018585777944988675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,128,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,256,0.008764444126023186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,256,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,512,0.017464889420403373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,64,0.004592888885074192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,32,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,128,0.008752889103359645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,128,0.00999644481473499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,256,0.014841778410805596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,1024,0.010663111176755695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,64,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,128,0.012080888781282636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,32,0.010006222460005019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,65536,0.20989955796135795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,65536,0.5042062335544163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,51200,0.34082667032877606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,16384,0.1313413381576538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,51200,0.1653617752922906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,65536,0.5879919793870714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,51200,0.45764711168077254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,12288,0.0975946651564704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,16384,0.05191377798716227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,65536,0.6048880153232151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,51200,0.4736986690097385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,10240,0.08234222067726983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,12288,0.040673779116736516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,12288,0.10271555847591823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,16384,0.1687093310885959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,8192,0.06654310888714261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,10240,0.0359173317750295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,10240,0.08984800179799397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,12288,0.12247378296322292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,7168,0.0564604467815823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,8192,0.028821332587136164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,10240,0.10451200273301865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,6144,0.04731111062897576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,7168,0.026334222820070054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,8192,0.08556533522076076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,7168,0.06276444594065349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,5120,0.0377146667904324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,512,0.011389333340856763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,6144,0.023853333459960088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,7168,0.07593511210547553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,6144,0.0551146666208903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,16384,0.13431999418470594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,5120,0.021002666817771062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,5120,0.04683644572893778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,6144,0.0663715534740024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,3584,0.027995556592941284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,4096,0.01793599956565433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,4096,0.039503110779656306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,5120,0.05709155400594076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,3072,0.02380977736579047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,4096,0.048582222726609975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,3584,0.016536000702116225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,3584,0.03519644339879354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,2560,0.020774222082561918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,8192,0.07147110833062066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,3072,0.015336889359686108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,3584,0.04461422231462267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,3072,0.0314355558819241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,2048,0.016460445192125108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,2560,0.01407911048995124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,2560,0.027169777287377253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,3072,0.039480000734329224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,1536,0.012757333616415659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,2048,0.012801777985360889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,2048,0.022800000177489385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,2560,0.03434933225313822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,1024,0.009454222189055549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,4096,0.0322746667597029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,2048,0.03022222386466132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,768,0.007822222179836696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,1024,0.01034222212102678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,1536,0.02497244377930959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,768,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,768,0.009716444545321995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,512,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,768,0.018201778332392376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,512,0.01075466639465756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,256,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,512,0.009101333717505137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,512,0.016812443733215332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,256,0.009722666607962714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,256,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,128,0.004559999952713649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,1536,0.01146044416560067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,1536,0.01959911154376136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,64,0.0041777777175108595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,256,0.014093332820468478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,32,0.004239999999602635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,128,0.009499555660618676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,1024,0.015118221441904703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,128,0.01107911103301578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,64,0.009319110876984065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,1024,0.020281778441535104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,32,0.009311999711725447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,65536,0.3797155486212836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,51200,0.24446845054626465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,65536,0.16252266036139593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,16384,0.08346666892369588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,65536,0.4961537785000271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,51200,0.3860168986850315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,16384,0.04183200001716614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,16384,0.11661867300669353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,65536,0.46715466181437176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,12288,0.07065511412090726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,10240,0.05563022030724419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,51200,0.36713867717319065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,12288,0.032251556714375816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,16384,0.12521421909332275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,128,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,12288,0.08983555767271255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,8192,0.047825776868396334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,10240,0.029947555727428857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,7168,0.04387022058169047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,10240,0.07566577858395047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,12288,0.09614311324225532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,8192,0.02350666622320811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,6144,0.03625511129697164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,8192,0.06207555532455444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,10240,0.08276000287797716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,7168,0.02132622235351139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,5120,0.03088088830312093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,7168,0.05636622508366903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,8192,0.06945777601665921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,6144,0.019261333677503798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,4096,0.024951110283533733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,6144,0.0483493341339959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,51200,0.12963822152879503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,7168,0.06111466884613037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,5120,0.01716977854569753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,3584,0.02180266711446974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,5120,0.04096444447835287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,6144,0.054392887486351855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,4096,0.014837332897716098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,3072,0.019352000620630052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,4096,0.03490133417977227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,5120,0.04634311132960849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,2560,0.016536888149049547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,3584,0.0310791101720598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,3072,0.0129457778400845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,4096,0.03980977667702569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,3072,0.02725155486000909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,2048,0.014207111464606391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,3584,0.03605866763326857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,2560,0.023927999867333308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,2560,0.011794666449228922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,3072,0.03273422188229031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,1536,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,2048,0.010671111444632212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,2048,0.02001333402262794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,2560,0.028806222809685603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,1024,0.008632889224423302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,1536,0.009883555273214975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,2048,0.025815111067559984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,1536,0.01686044368478987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,768,0.0075839997993575195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,1024,0.0092613332801395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,1536,0.02128355536195967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,512,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,768,0.011620444556077322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,1024,0.01757066614098019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,768,0.016153777639071148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,3584,0.014078223043017916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,512,0.009718221922715506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,256,0.0047155556579430895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,256,0.007738666401969061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,512,0.014496889379289416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,256,0.009029332962301042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,256,0.013099555340078143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,128,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,64,0.003547555456558863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,128,0.007932444413503012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,32,0.0038399998512532976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,128,0.01039644413524204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,1024,0.013410666750537025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,64,0.00870577742656072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,32,0.008660444782839881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,768,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,65536,0.3523111078474257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,512,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,51200,0.23334044880337187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,65536,0.43179821968078613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,128,0.003908444609906939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,16384,0.08800266848670112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,51200,0.12088799476623535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,51200,0.3537306785583496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,12288,0.06711733341217041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,16384,0.1026088860299852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,16384,0.04052799940109253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,51200,0.3613128927018907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,12288,0.03199822372860379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,10240,0.058054222000969775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,12288,0.07918222082985772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,10240,0.028734223710166082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,12288,0.0948408908314175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,8192,0.044701334502961904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,10240,0.06685600015852186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,65536,0.15402132934994167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,7168,0.03841955463091532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,10240,0.08109066883722942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,8192,0.023016000787417095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,8192,0.05441955394215054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,6144,0.03471644388304816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,7168,0.021031111478805542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,8192,0.06694577799903022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,7168,0.04849511053827074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,5120,0.02850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,65536,0.4568302366468642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,6144,0.018760000665982563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,6144,0.04264533188607958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,7168,0.059417777591281466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,5120,0.01682222220632765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,6144,0.053096890449523926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,3584,0.021495999561415777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,16384,0.1241884496476915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,4096,0.014821334017647637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,4096,0.03107200066248576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,3072,0.01878666712178124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,5120,0.04613955484496223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,3584,0.013783999615245394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,4096,0.03874933388498094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,2560,0.01657511128319634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,3584,0.02779200010829502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,3072,0.012712000144852532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,3072,0.02463022205564711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,3584,0.03549777799182468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,2048,0.013907555076811047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,2560,0.011710222396585675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,2560,0.021706667211320665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,1536,0.011179555621412067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,2048,0.010487111078368293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,2560,0.02823466724819607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,1536,0.01517777807182736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,4096,0.024545777175161574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,1536,0.009657777845859528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,5120,0.036899556716283165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,1536,0.02094488839308421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,1024,0.012352888782819113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,1024,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,1024,0.017094223035706412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,768,0.010571555958853828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,512,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,768,0.008631111019187504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,768,0.015839111473825242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,3072,0.03180888957447476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,512,0.00903733323017756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,512,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,256,0.004713777866628435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,2048,0.01849688920709822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,512,0.013949332965744866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,2048,0.02427822185887231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,256,0.007683555285135905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,1024,0.008229333493444655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,256,0.008290666672918532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,64,0.0038319999972979226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,128,0.007717333734035492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,32,0.004230222354332606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,256,0.012799111505349478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,768,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,64,0.007668444679843054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,128,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,32,0.007704888780911763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,51200,0.191901339424981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,65536,0.3389431105719672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,128,0.003964444415436851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,65536,0.4564204216003418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,16384,0.06547733147939046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,128,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,51200,0.11592177549997966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,51200,0.3579564359452989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,12288,0.04967733224232992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,16384,0.038204444779290095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,65536,0.25228799713982475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,16384,0.08597866694132487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,10240,0.04182044333881802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,12288,0.030827555391523574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,65536,0.1452595525317722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,16384,0.12275555398729111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,12288,0.06472088893254598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,8192,0.030701332622104224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,10240,0.027453333139419556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,10240,0.0554986662334866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,7168,0.02767555581198798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,12288,0.09357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,8192,0.02214844524860382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,8192,0.04558666547139486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,6144,0.023137778043746948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,51200,0.26613156000773114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,7168,0.01979200045267741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,10240,0.07981510957082112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,7168,0.040525333748923406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,5120,0.020564445190959506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,8192,0.06643911202748616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,6144,0.01812977757718828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,4096,0.01628533336851332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,6144,0.03575999869240655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,5120,0.015995555453830294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,7168,0.059333331055111356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,3584,0.015012444721327888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,5120,0.030290666553709242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,4096,0.014192889134089151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,6144,0.052575111389160156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,3072,0.01331822243001726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,4096,0.02554133368863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,3584,0.013255111045307584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,5120,0.045475555790795215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,2560,0.011419555379284753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,3584,0.023399111297395494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,3072,0.012138666378127204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,4096,0.03767377800411648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,2048,0.009323555562231276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,3072,0.020392000675201416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,3584,0.03463022245301141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,2560,0.01789066692193349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,1536,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,3072,0.030744890371958416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,2048,0.010393777655230628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,2048,0.015586665934986539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,1024,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,2560,0.02709688908523983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,1536,0.00941244430012173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,1536,0.013236444857385425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,2048,0.023545778459972803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,1024,0.008980444735950893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,1024,0.010678222609890832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,512,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,1536,0.02052177819940779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,768,0.008626666333940294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,1024,0.017190222938855488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,768,0.009106666677527958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,256,0.0036115555299652945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,512,0.008238222036096785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,768,0.01517955627706316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,512,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,128,0.0031724443866146933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,256,0.007347555624114142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,256,0.007735999921957652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,64,0.0028817777832349143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,512,0.013788444300492605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,32,0.0031848889258172777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,128,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,128,0.007739555504586961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,2560,0.011193777951929303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,64,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,65536,0.16995644569396973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,32,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,768,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,65536,0.10002399815453424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,51200,0.13732533984714085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,65536,0.28183555603027344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,16384,0.04722399844063652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,51200,0.07989688714345296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,65536,0.31421422958374023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,51200,0.22251823213365343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,16384,0.07259733147091336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,16384,0.02698488864633772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,12288,0.03581155671013726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,51200,0.2461564540863037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,12288,0.022389334109094407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,10240,0.02986577815479702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,256,0.01276088919904497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,16384,0.08752444717619155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,12288,0.055346667766571045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,8192,0.02342844506104787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,128,0.010124444133705562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,12288,0.0673440032535129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,10240,0.02039377722475264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,10240,0.046909332275390625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,7168,0.02107111116250356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,8192,0.016081778539551627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,8192,0.03863644599914551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,6144,0.018400000201331247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,10240,0.05779111385345459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,7168,0.034770667552948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,5120,0.01628977722591824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,8192,0.048114667336146034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,6144,0.013428444663683573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,6144,0.03053599927160475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,7168,0.04414133230845133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,4096,0.013555555707878537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,5120,0.012233778006500669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,6144,0.03843377696143256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,5120,0.026141333911154006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,3584,0.012283555335468717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,4096,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,4096,0.022078222698635522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,3072,0.010774222513039907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,5120,0.033364444971084595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,3584,0.01057866629627016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,3584,0.019967110620604623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,2560,0.0094035557574696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,3072,0.010055111514197456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,3072,0.018176888426144917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,3584,0.02586577832698822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,2048,0.007727999654081132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,2560,0.016030222177505493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,2560,0.00941955546538035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,3072,0.02348088887002733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,1536,0.006540444576077991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,7168,0.014646222194035849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,2048,0.00903200027015474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,1024,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,1536,0.008319999608728621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,2560,0.020977778567208186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,1536,0.011720000041855706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,768,0.004761777818202972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,2048,0.018925334016482036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,1536,0.016493333710564505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,1024,0.009529777699046666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,512,0.0038888889054457345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,768,0.007786666353543599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,768,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,1024,0.014154665999942355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,256,0.0033244445092148254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,4096,0.028205331828859117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,512,0.007341333561473423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,512,0.0070062221752272705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,768,0.013103110922707452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,128,0.003134222287270758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,256,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,512,0.01182755579551061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,64,0.00281866660548581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,128,0.006311111152172089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,128,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,32,0.0029164445069101122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,256,0.01130222198035982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,2048,0.013797333670987023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,128,0.008714666797055138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,65536,0.14592266082763672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,64,0.006435555716355641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,32,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,51200,0.11614755789438884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,1024,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,65536,0.09901422262191772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,65536,0.23982667922973633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,51200,0.077583114306132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,51200,0.18877689043680826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,12288,0.02780710988574558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,16384,0.027154665854242113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,16384,0.062436441580454506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,65536,0.3096008830600315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,10240,0.023306666149033442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,51200,0.24563821156819662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,12288,0.021703999903466966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,256,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,16384,0.0860862202114529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,12288,0.048342221313052706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,8192,0.018772444791264005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,10240,0.019865777757432725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,10240,0.0409786668088701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,12288,0.06654933426115248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,8192,0.015726221932305228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,8192,0.0340648889541626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,10240,0.05754577451282077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,6144,0.015166223049163818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,8192,0.048023998737335205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,7168,0.030552887254291113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,16384,0.03892711136076186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,5120,0.013341333303186627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,6144,0.013460444079505073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,6144,0.026909333136346605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,7168,0.04285155402289497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,4096,0.011015999648306104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,5120,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,6144,0.03788355655140347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,5120,0.023183999790085688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,3584,0.010006222460005019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,4096,0.011085333095656501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,5120,0.03312799996799893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,4096,0.019510222805870902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,3072,0.009002666506502364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,3584,0.017809778451919556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,3584,0.010426667001512317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,4096,0.02784088916248745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,7168,0.016912889149453905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,2560,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,3072,0.00979911122057173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,3072,0.016100444727473788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,3584,0.02532622218132019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,2048,0.006729777902364731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,2560,0.01424888935354021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,2560,0.009704888694816166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,1536,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,3072,0.023356444305843774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,2560,0.020617778102556866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,2048,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,2048,0.01241422196229299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,1024,0.005195555587609609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,1536,0.008398222426573435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,1536,0.010470222267839642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,2048,0.01867644488811493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,768,0.004554666578769684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,1024,0.007778666913509369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,1024,0.00885511106914944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,1536,0.01620444489849938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,512,0.0038240001433425476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,1024,0.013781333135233985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,768,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,256,0.003176888864901331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,768,0.013095999757448832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,512,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,128,0.0028328889360030494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,256,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,256,0.006231110956933763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,7168,0.014556444353527494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,128,0.005969777703285217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,64,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,256,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,128,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,65536,0.10362311204274495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,128,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,64,0.005966222120655908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,32,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,51200,0.08022133509318034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,65536,0.09763289160198635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,65536,0.2004906601376004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,768,0.007684444387753804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,16384,0.03243466549449497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,512,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,51200,0.16018933720058864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,65536,0.29764800601535374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,512,0.011766222616036734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,16384,0.05287466777695549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,16384,0.02574311031235589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,51200,0.2362453407711453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,10240,0.019654222660594516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,12288,0.020799110333124798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,16384,0.08186311191982694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,10240,0.01899111105336083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,12288,0.06435022089216444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,10240,0.03530399998029073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,8192,0.016183111402723525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,10240,0.05498488744099935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,8192,0.015169777803950839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,7168,0.014814222852389017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,8192,0.029113776153988306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,7168,0.014153778553009033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,7168,0.026226666238572862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,51200,0.07899644639756945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,6144,0.01310222182008955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,12288,0.02297244469324748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,6144,0.013064000341627332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,7168,0.041839109526740186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,6144,0.02315644423166911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,32,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,12288,0.04124622212515937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,5120,0.011715555356608497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,5120,0.020261334048377145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,6144,0.03606222073237101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,4096,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,3584,0.008657777474986183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,5120,0.03130666746033563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,4096,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,4096,0.01715377800994449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,3072,0.007706666986147563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,3584,0.015796444482273526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,3584,0.010065777434243096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,4096,0.026935999592145283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,2560,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,3072,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,3072,0.01408088869518704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,8192,0.045949333243899874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,2048,0.006211555666393704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,2560,0.009312888814343346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,3072,0.022286223040686712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,5120,0.01127199994193183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,1536,0.005315555466545953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,2048,0.009002666506502364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,2048,0.011038222246699862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,2560,0.020271999968422782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,1024,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,1536,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,2048,0.017678222722477384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,1536,0.009617778162161509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,768,0.003906666818592283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,1024,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,1536,0.015776000089115567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,768,0.007128000259399414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,768,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,512,0.0035253332720862497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,1024,0.013691555294725629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,256,0.0028400001012616684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,3584,0.02475555572244856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,512,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,768,0.01240355521440506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,512,0.00655733338660664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,128,0.0028248888750871024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,2560,0.012697777814335294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,256,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,64,0.0024862223201327855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,128,0.0069848886794514125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,128,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,32,0.002557333351837264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,256,0.01073688848151101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,65536,0.05863644679387411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,128,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,32,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,51200,0.04768088791105482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,1024,0.008021333151393467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,65536,0.09724533557891846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,65536,0.15754932827419704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,16384,0.017652443713612027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,51200,0.07832888762156169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,51200,0.12365510728624131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,65536,0.2939528889126248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,16384,0.042261332273483276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,12288,0.01702311138312022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,16384,0.024653333756658766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,51200,0.2324168947007921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,512,0.011399110986126794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,10240,0.015279110934999255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,12288,0.033032887511783175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,16384,0.08049688736597697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,8192,0.012679999901188744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,10240,0.018969777557584975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,12288,0.06295466423034668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,64,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,10240,0.028350220786200628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,7168,0.011476444701353708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,8192,0.01516355574131012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,8192,0.02370311154259576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,6144,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,10240,0.05435644586881002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,7168,0.02125599980354309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,7168,0.014008000493049622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,5120,0.00868622213602066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,8192,0.04528177777926127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,6144,0.0127688886390792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,7168,0.04080444574356079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,6144,0.018925334016482036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,4096,0.007289778027269576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,5120,0.011749332977665795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,5120,0.016893333858913846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,6144,0.035338666703965925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,3584,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,256,0.006870222174459034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,4096,0.014451555079884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,4096,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,3072,0.006137777947717243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,5120,0.03047555685043335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,3584,0.010058666268984476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,4096,0.026404443714353774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,3584,0.013321778012646569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,3072,0.009703111317422655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,3072,0.012091555529170565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,3584,0.023987554841571387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,2560,0.009071110851234859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,2048,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,3072,0.02167466613981459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,2048,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,2560,0.019952888290087383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,1536,0.003906666818592283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,2048,0.01758044461409251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,1536,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,1536,0.008269333177142674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,1024,0.0032666667054096856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,1536,0.015447111593352424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,12288,0.020709334148301017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,1024,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,1024,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,768,0.0031902222997612427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,2560,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,512,0.00282222218811512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,768,0.007522666619883643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,768,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,2560,0.011035555766688453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,512,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,256,0.002471999989615546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,768,0.012070222033394707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,512,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,128,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,2048,0.009719111025333405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,256,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,512,0.011383111278216044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,256,0.005281777845488654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,64,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,32,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,128,0.005061333378156026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,128,0.006676444576846228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,256,0.010631110933091907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,65536,0.04448622133996752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,64,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,128,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,51200,0.03552622265285916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,32,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,65536,0.09717688957850139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,16384,0.011119999819331698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,65536,0.1565706729888916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,51200,0.07787999841901991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,51200,0.12300977441999648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,1024,0.013212444053755866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,12288,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,16384,0.041497776905695595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,65536,0.29323111640082467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,10240,0.010784000158309937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,51200,0.23214756117926705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,12288,0.02033511135313246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,16384,0.08097422122955322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,12288,0.03251555562019348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,8192,0.009713778065310584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,10240,0.018927999668651156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,10240,0.027722666660944622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,7168,0.01072355525361167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,12288,0.06268622477849324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,8192,0.02311022248533037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,10240,0.053328000836902194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,6144,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,7168,0.013917333549923368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,7168,0.020920889245139226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,8192,0.044123556878831655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,5120,0.008272888759771982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,6144,0.012794666820102267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,6144,0.018889778190188937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,7168,0.03906666570239597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,4096,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,5120,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,5120,0.016540444559521146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,6144,0.035028444396124944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,3584,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,16384,0.02414488957987891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,4096,0.014350222216712104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,5120,0.03051911128891839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,3072,0.005676444619894028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,3584,0.010008888940016428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,3584,0.01313244468635983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,4096,0.02643999954064687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,2560,0.005127111242877113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,3072,0.012111110819710625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,3584,0.02405244443151686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,8192,0.014997333288192749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,2048,0.004517333375083076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,3072,0.02167466613981459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,2560,0.01090755561987559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,2560,0.00925422211488088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,1536,0.0038488888078265717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,2048,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,2560,0.019458666443824768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,2048,0.009666666388511658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,1024,0.0032177778581778207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,1536,0.008542222281297049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,1536,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,768,0.0028640000770489373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,2048,0.017601778109868366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,1024,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,1024,0.007639110916190677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,1536,0.015531554818153381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,512,0.0025680000997251934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,4096,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,1024,0.013024888932704926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,768,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,256,0.0021991110924217436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,512,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,512,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,768,0.011755555868148804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,128,0.0021137777302000257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,3072,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,256,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,512,0.010781333678298526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,64,0.0018835556175973679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,128,0.00675999994079272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,128,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,32,0.002143111079931259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,256,0.009739555418491364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,65536,0.04182577795452542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,51200,0.03319999906751845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,64,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,16384,0.013635555903116861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,32,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,12288,0.01146755533085929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,65536,0.15625777509477404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,10240,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,51200,0.12257511085934109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,768,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,8192,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,16384,0.04140355520778232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,7168,0.008926221893893348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,12288,0.03224444389343262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,6144,0.00795199970404307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,5120,0.007332444190979004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,10240,0.027383999692069158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,4096,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,8192,0.023057777020666335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,3584,0.006220444622966979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,256,0.006703111032644908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,3072,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,6144,0.018761777215533786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,2560,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,5120,0.016542222764756944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,4096,0.014392000105645923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,3584,0.013074666261672974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,1536,0.0038355555799272326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,3072,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,1024,0.003142222141226133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,768,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,2560,0.010949333508809408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,2048,0.00960533320903778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,512,0.0025395556456512874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,256,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,1536,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,128,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,1024,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,64,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,512,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,32,0.001962666710217794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,256,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,65536,0.045179555813471474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,64,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,51200,0.033801777495278254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,32,0.004944889081848992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,16384,0.009319110876984065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,7168,0.020971554848882887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,12288,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,65536,0.1552657816145155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,2048,0.004567111117972268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,10240,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,8192,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,51200,0.1225279966990153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,16384,0.041293332974116005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,7168,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,6144,0.00908799966176351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,12288,0.03192355566554599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,5120,0.008588444855478074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,10240,0.02754844393995073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,4096,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,8192,0.023036445180575054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,3584,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,7168,0.020982222424613103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,3072,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,768,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,5120,0.016576000385814242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,2560,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,2048,0.005143110950787862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,4096,0.014409777190950183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,1536,0.0036026665733920205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,3584,0.01312711089849472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,3072,0.012065778175989786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,1024,0.0031537777847713898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,768,0.0028115556471877625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,2560,0.010953777366214328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,2048,0.009541333549552495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,512,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,256,0.0021893332401911416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,1536,0.008319999608728621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,128,0.001995555642578337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,1024,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,768,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,512,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,64,0.001814222170246972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,32,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,256,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,128,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,64,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,32,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,6144,0.01871200071440803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,128,0.005038222091065513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,16384,0.6859404246012369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,16384,2.029559029473199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,16384,2.2328328026665583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,12288,0.5133173200819228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,12288,1.7480986913045247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,51200,2.5296942392985025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,12288,1.5938427183363173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,10240,1.4334772957695856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,10240,0.43146223492092556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,10240,1.4141564899020727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,8192,0.3768151071336534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,8192,1.0371821721394856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,8192,1.3403102027045355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,7168,0.3216408888498942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,7168,1.2998249265882704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,7168,0.8920044369167752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,6144,0.2759137683444553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,6144,0.8359804683261447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,51200,6.454594506157769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,5120,0.23511110411749947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,6144,0.9897698296440972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,5120,0.8989440070258247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,4096,0.5757173432244195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,5120,0.6210026741027832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,4096,0.19846310880449083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,3584,0.5434977743360732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,3584,0.1744204494688246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,3072,0.39095022943284774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,4096,0.5289582146538628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,3584,0.47700177298651797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,3072,0.3923644489712185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,3072,0.19831821653578016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,2560,0.35372622807820636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,2560,0.13450133800506592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,2048,0.3139128949907091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,2048,0.12090844578213161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,2560,0.3312275674608019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,2048,0.26488534609476727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,1536,0.09563999705844456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,1024,0.16142222616407606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,1536,0.2066826687918769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,768,0.13400977187686497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,1024,0.07208622164196439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,51200,8.584540473090277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,768,0.06404444244172838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,1024,0.15690222051408556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,768,0.13014400005340576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,256,0.054888000090916954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,128,0.04167911079194811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,512,0.06040088997946846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,256,0.05566755268308851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,512,0.10866222116682266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,256,0.09707555505964492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,32,0.04331022169854906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,128,0.09153155485788982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,64,0.09117155604892307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,1536,0.22534134652879503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,32,0.09211733606126572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,512,0.08061333497365315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,64,0.04115466607941522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,128,0.053054223457972206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,65536,2.427489810519748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,16384,1.551571528116862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,51200,1.9480533599853516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,51200,4.701272752549913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,16384,0.5309013260735406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,12288,1.3694178263346355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,12288,0.4340026643541124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,65536,6.1725209554036455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,16384,1.880489773220486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,10240,1.1465750800238717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,12288,1.247911983066135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,8192,0.9684444003634982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,10240,0.34315911928812665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,10240,1.0900150934855144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,8192,0.283315552605523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,51200,6.731693267822266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,7168,0.24860265519883898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,8192,0.849147531721327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,7168,0.7978737619188098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,6144,0.6646106508043077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,7168,0.7149146927727593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,6144,0.21663378344641793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,5120,0.5453226831224229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,5120,0.2153182162178887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,4096,0.3820915487077501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,65536,8.766186184353298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,4096,0.4076480070749919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,4096,0.17600977420806885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,3584,0.44200799200269913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,3584,0.13752888308631048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,3584,0.3597875436147054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,3072,0.3412204318576389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,3072,0.12219022379981147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,3072,0.3455893463558621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,2560,0.2448071108924018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,6144,0.6255288653903538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,2048,0.2174604468875461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,2560,0.1081084476576911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,2048,0.09109777874416775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,1536,0.15538756052652994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,2048,0.22288266817728677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,1536,0.07200088765886095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,1024,0.10578133662541707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,128,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,1536,0.16311644183264837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,768,0.09981599781248306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,1024,0.05697421895133125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,512,0.0758168896039327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,1024,0.12586932712131077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,768,0.051580445633994214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,256,0.04996977912055122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,512,0.052143110169304736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,512,0.08637333578533596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,128,0.037802666425704956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,256,0.04410755634307861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,256,0.07704444726308186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,64,0.031011554929945204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,32,0.03240799903869629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,128,0.04285866684383816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,128,0.07287022140291002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,2560,0.2990115483601888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,32,0.07338311274846394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,64,0.07248266537984212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,5120,0.5159795549180772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,768,0.09925244251887004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,16384,0.4497377607557509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,65536,0.7529386414421929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,51200,0.59660890367296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,16384,0.534695095486111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,16384,0.18299289544423422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,12288,0.4359466764662001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,51200,1.5846053229437933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,12288,0.14005688826243082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,10240,0.3146897686852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,65536,2.185624016655816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,12288,0.43984532356262207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,10240,0.11783111095428467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,8192,0.2820613384246826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,10240,0.4502711296081543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,8192,0.1125662194357978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,7168,0.20401067203945586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,8192,0.270576900906033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,7168,0.08566221925947402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,6144,0.199492441283332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,5120,0.14674311214023167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,7168,0.22859644889831543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,6144,0.07474755578570895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,5120,0.07206310828526814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,6144,0.2269244458940294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,5120,0.17530755201975504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,4096,0.05370044377115038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,3584,0.1381093396080865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,4096,0.13401510980394152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,3072,0.11149956120385064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,3584,0.04826844400829739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,2560,0.0853946672545539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,3584,0.11767555607689752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,3072,0.043155554268095225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,2048,0.0737982193628947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,3072,0.10636622375912136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,2560,0.03811644514401754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,1536,0.049504889382256397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,2560,0.08835289213392471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,2048,0.03282399972279867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,51200,1.9985635545518663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,1024,0.03614755471547445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,2048,0.07327289051479764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,1536,0.027265777190526325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,1536,0.05771288606855604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,512,0.021764443980322942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,1024,0.04427644279268053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,768,0.020257777637905545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,65536,2.7701085408528647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,256,0.015951999359660678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,768,0.036044445302751325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,128,0.01204711116022534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,4096,0.13706578148735896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,512,0.018400000201331247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,64,0.01072533345884747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,256,0.016606221596399944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,512,0.029507554239696924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,32,0.010695999695195092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,128,0.01685066686736213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,256,0.026064000195927087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,128,0.025460445218616064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,64,0.025119110941886902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,32,0.025090666280852422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,768,0.027973333994547527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,1024,0.022175111704402502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,65536,0.5851422415839301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,16384,0.4238408936394586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,51200,1.248662206861708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,16384,0.1398577822579278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,65536,1.7432550854153102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,16384,0.48594220479329425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,12288,0.3189280033111572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,51200,1.5040302276611328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,12288,0.11661956045362686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,65536,1.8801449669731989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,10240,0.28265510665045845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,10240,0.09364000293943618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,8192,0.220268448193868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,10240,0.26935378710428876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,51200,0.4633964432610406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,7168,0.18885244263543022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,8192,0.09005955855051677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,6144,0.16205067104763454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,8192,0.20668799347347686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,6144,0.0631093316608005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,5120,0.1347422202428182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,7168,0.18681244055430093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,6144,0.15622044934166804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,5120,0.05181688732571072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,3584,0.09873955779605442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,4096,0.04201333390341865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,4096,0.10415466626485188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,3584,0.03779822256830003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,3072,0.08060710959964328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,12288,0.30607467227511936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,3584,0.09184355868233575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,2560,0.06729333268271552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,3072,0.03415022293726603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,2048,0.05481511023309496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,3072,0.0797662205166287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,2560,0.03011111087269253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,1536,0.038819554779264666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,2560,0.06852622164620294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,1024,0.02790400054719713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,2048,0.025978666212823655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,7168,0.06668621963924833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,1536,0.021984888447655573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,2048,0.05714755588107639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,768,0.022433777650197346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,512,0.016941333810488384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,1536,0.04514400164286295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,1024,0.018400000201331247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,256,0.012504888905419244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,768,0.016864000095261466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,4096,0.10641244384977554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,512,0.015279110934999255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,768,0.028976887464523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,5120,0.12624977694617376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,512,0.02346933384736379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,256,0.014139556222491793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,64,0.008279999925030602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,32,0.008348444269763099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,256,0.02069422271516588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,128,0.02030755579471588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,64,0.02032977839310964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,32,0.020301333732075162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,65536,0.48328532112969297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,1024,0.03457955519358317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,128,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,51200,0.8254204326205783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,65536,1.3297938240898979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,128,0.014335110783576965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,16384,0.33707287576463485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,16384,0.1233324474758572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,51200,0.38086843490600586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,12288,0.2746693293253581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,16384,0.34022755093044704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,65536,1.6406151453653972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,10240,0.20893422762552896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,12288,0.24989954630533853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,8192,0.15421422322591147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,10240,0.09502133395936753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,7168,0.13688800070020887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,8192,0.06759822368621826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,10240,0.275036440955268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,6144,0.12912889321645102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,7168,0.06083199712965223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,8192,0.17949510945214164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,5120,0.1031422217686971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,7168,0.1534631119834052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,6144,0.05010222064124214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,4096,0.0829368895954556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,5120,0.044362667534086436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,6144,0.13266844219631618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,3584,0.07005066341824003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,4096,0.03616355525122748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,5120,0.11515022648705377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,3072,0.06348088714811537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,3584,0.03240711159176297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,4096,0.08881866931915283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,2560,0.05468177795410156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,3072,0.029313776228162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,3584,0.07817333274417453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,2560,0.026608000199000042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,3072,0.06860711177190144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,12288,0.0926275584432814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,2048,0.022854222191704646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,2560,0.058965331978268094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,51200,1.1705599890814888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,1024,0.02482755482196808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,2048,0.0491866668065389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,768,0.01862844493654039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,1536,0.019455111689037748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,1024,0.01650311052799225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,1536,0.038903110557132296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,512,0.014280888769361707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,1024,0.03002133303218418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,768,0.014833778142929077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,256,0.010372444159454769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,768,0.024873778223991394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,128,0.0088746663596895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,512,0.020346666375796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,256,0.012763555679056378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,64,0.007481777833567725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,256,0.018208889497650992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,32,0.00739644467830658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,128,0.017620444297790527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,2048,0.04582755433188545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,64,0.017620444297790527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,32,0.017607111069891188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,1536,0.03257155418395996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,65536,0.41732178794013125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,512,0.01369955556260215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,51200,0.7415866851806641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,65536,1.1002800199720595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,128,0.01277599980433782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,16384,0.27548710505167645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,12288,0.21227733294169107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,16384,0.1316524479124281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,51200,0.9603964487711588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,51200,0.32874756389194065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,65536,1.3588541878594294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,16384,0.3619102107153998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,12288,0.20411022504170737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,10240,0.20715110831790498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,10240,0.07192444139056735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,8192,0.13089956177605525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,7168,0.12408978409237331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,8192,0.053232000933753125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,6144,0.09852088822258843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,10240,0.1714177793926663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,7168,0.04987644486957126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,5120,0.09648355510499741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,6144,0.041658666398790144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,4096,0.07648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,7168,0.1215297778447469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,5120,0.03664888938268026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,6144,0.1179013384713067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,3584,0.06266933017306857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,5120,0.08911999728944565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,3072,0.05375911129845513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,3584,0.030573334958818223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,2560,0.04838666650984022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,3072,0.024832889437675476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,3584,0.06401866674423218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,3072,0.055973331133524575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,2048,0.03735555542839898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,2560,0.022595556245909795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,1536,0.02703199949529436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,2560,0.049514667855368726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,12288,0.07851999998092651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,1024,0.019388443893856473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,2048,0.04077333211898804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,1536,0.017421333326233756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,768,0.016167999969588388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,1536,0.03301066822475857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,1024,0.014310222533014087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,512,0.012335999972290464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,768,0.013088000317414602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,768,0.0214035560687383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,256,0.008940444224410588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,512,0.012183999849690331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,4096,0.035818666219711304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,512,0.017440888616773818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,128,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,4096,0.07270399729410808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,256,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,64,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,32,0.0064479998416370815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,256,0.015511110424995422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,128,0.011368888947698804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,128,0.01519555515713162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,64,0.0151573336786694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,32,0.01514844430817498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,2048,0.019928000039524503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,8192,0.14178755548265246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,1024,0.025359110699759588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,65536,0.36617599593268496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,51200,0.2905457814534505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,16384,0.23551289240519205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,65536,1.0417306688096788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,51200,0.7529306411743164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,16384,0.09155199925104777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,12288,0.18877066506279838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,16384,0.245140446556939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,10240,0.20867910650041369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,12288,0.08770489030414158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,51200,0.8682808876037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,65536,1.2144720289442275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,12288,0.1836657789018419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,8192,0.16502311494615343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,7168,0.12560444407992893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,8192,0.050276445017920605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,6144,0.10806044605043198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,8192,0.1336853371726142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,7168,0.05780800183614095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,7168,0.10965955257415771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,5120,0.0890844464302063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,6144,0.044600887431038745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,4096,0.07001511255900066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,6144,0.09781599707073635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,5120,0.03291822142071194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,3584,0.05978666411505806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,5120,0.08657688564724392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,4096,0.0284204449918535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,3072,0.04915822214550442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,4096,0.06722844309277005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,3584,0.026124444272783067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,2560,0.03928000066015456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,3584,0.05738933218850029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,3072,0.023295111126369897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,2048,0.030800888935724895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,3072,0.05059555504057142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,2560,0.02197244432237413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,1536,0.022689777943823073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,2560,0.043769776821136475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,1024,0.016520889268981088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,10240,0.06149599949518839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,10240,0.1570568879445394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,1536,0.01587377819750044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,768,0.01386755539311303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,1536,0.029884444342719182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,1024,0.013459555804729462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,512,0.011122666299343109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,1024,0.023066666391160753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,768,0.012794666820102267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,512,0.011699555648697747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,768,0.019351111518012155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,128,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,512,0.015836444165971544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,256,0.010855110983053843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,64,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,256,0.014369777507252164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,128,0.011090666883521609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,32,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,128,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,64,0.013786666923099093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,32,0.013747555514176687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,2048,0.018529777725537617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,2048,0.03751022285885281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,65536,0.6726080046759711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,65536,0.3422559897104899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,16384,0.2113920052846273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,256,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,51200,0.5202862421671549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,51200,0.26977332433064777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,12288,0.15535822179582384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,16384,0.08520622385872735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,65536,0.9528587129380969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,12288,0.07162311342027453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,10240,0.1333857774734497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,10240,0.05572355455822415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,12288,0.15811378426022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,8192,0.10565066337585449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,51200,0.7269137700398763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,7168,0.09595377577675714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,10240,0.13900444242689344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,8192,0.04214310977194044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,7168,0.04188444548183017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,6144,0.08120089107089572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,5120,0.0676435563299391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,8192,0.11239911450280084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,6144,0.034104890293545194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,5120,0.030361778206295435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,6144,0.08574755324257745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,3584,0.04713866776890225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,4096,0.025496888491842482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,3072,0.03970044520166185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,3584,0.023494222097926672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,3584,0.0518506666024526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,2560,0.0310533344745636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,3072,0.045215109984079994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,2560,0.01906755566596985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,16384,0.25254578060574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,2560,0.039670222335391574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,2048,0.016691555579503376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,1536,0.020248000820477802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,2048,0.033278223541047834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,7168,0.09611466858122085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,1024,0.015223110715548197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,1536,0.014567111929257711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,5120,0.072525335682763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,1536,0.027219555444187585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,768,0.01275022245115704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,1024,0.012488889197508493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,4096,0.058283554183112256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,1024,0.0207368897067176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,512,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,768,0.011744889120260874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,256,0.007289778027269576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,768,0.017483555608325534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,3072,0.021309332715140447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,512,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,512,0.014504889647165934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,256,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,64,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,256,0.01311377767059538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,32,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,2048,0.02551555633544922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,128,0.010346666806273991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,128,0.01277955538696713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,64,0.012784889174832238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,32,0.012494222157531314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,4096,0.051797333690855235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,65536,0.266941335466173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,65536,0.6586862140231663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,51200,0.49835019641452366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,51200,0.21117599805196127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,16384,0.16514133082495794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,12288,0.12398489316304524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,128,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,65536,0.8181058035956489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,51200,0.6313778029547797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,12288,0.05146844519509209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,10240,0.1137288941277398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,12288,0.13694666491614446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,8192,0.08551466464996338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,10240,0.11650489436255561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,7168,0.07508355379104614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,8192,0.03616888986693488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,8192,0.0920284456676907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,6144,0.06009066767162747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,7168,0.032511999209721885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,5120,0.052588442961374916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,7168,0.08277689086066352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,6144,0.029117332564459905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,4096,0.04264800084961785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,6144,0.07078578074773152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,5120,0.025407999753952026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,5120,0.06113511323928833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,4096,0.021895999709765118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,3072,0.032030221488740705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,4096,0.0496417780717214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,16384,0.0798471106423272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,3584,0.044600887431038745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,2560,0.02611911131276025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,16384,0.21018844180636934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,3072,0.01826844447188907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,2048,0.021738666627142165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,3072,0.03940533267127143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,10240,0.046088887585533984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,2560,0.016696888539526198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,1536,0.01687733332316081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,1024,0.012654222548007965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,2560,0.034360888931486346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,2048,0.01477688882086012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,1536,0.013122667041089801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,2048,0.029334220621320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,1536,0.023361777265866596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,1024,0.011420444481902652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,512,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,768,0.010659555594126383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,768,0.015498666299713982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,512,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,512,0.01309333327743742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,128,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,256,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,256,0.011743110915025076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,64,0.005022222383154763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,128,0.009524444739023844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,32,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,128,0.011385777758227454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,64,0.011360888679822287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,32,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,768,0.010325333310498131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,65536,0.42073244518703884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,1024,0.018026666508780587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,65536,0.24113421969943574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,256,0.006215999937719769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,51200,0.41493066151936847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,16384,0.13093333774142796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,3584,0.039001779423819646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,65536,0.6522889137268066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,12288,0.10234310891893174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,16384,0.06830310821533203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,10240,0.09593333138359918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,3584,0.020170667105250888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,16384,0.14876088831159803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,51200,0.5322684182061089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,12288,0.050847109821107656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,8192,0.07538044452667236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,7168,0.06575644678539701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,12288,0.11315022574530707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,8192,0.03385333220163981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,6144,0.05436088972621494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,10240,0.10196889109081692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,7168,0.030815998713175457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,8192,0.07756355735990736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,5120,0.046367999580171376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,6144,0.028008888165156048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,6144,0.05992088715235392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,5120,0.02438755498992072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,5120,0.05081244309743246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,4096,0.020829333199395072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,4096,0.041787554820378624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,51200,0.19317155414157441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,3072,0.0270480000310474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,3584,0.01926933394538032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,3584,0.03791288865937127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,2560,0.023507555325826008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,3072,0.01755288905567593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,10240,0.04714488983154297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,3072,0.03349422083960639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,2048,0.019112888309690688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,2560,0.016027554869651794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,2560,0.02920711040496826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,1536,0.014949333336618213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,2048,0.0140657772620519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,1024,0.011379555695586734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,2048,0.024675556355052527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,1536,0.012440000143316058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,768,0.009292444421185387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,1536,0.019924443629052904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,1024,0.01090844472249349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,512,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,1024,0.01581333412064446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,768,0.01036622209681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,3584,0.02845777736769782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,256,0.005506666584147348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,768,0.013459555804729462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,128,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,512,0.009671111073758867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,256,0.009023110899660323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,512,0.011375111010339526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,64,0.0041777777175108595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,32,0.004511111312442356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,128,0.009296889106432596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,256,0.010488889283604093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,128,0.010066666536860997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,64,0.010048888623714447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,32,0.00999466660949919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,7168,0.06840266784032185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,65536,0.23388976520962187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,51200,0.3858577675289578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,4096,0.03700088792377048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,16384,0.11939999792310928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,65536,0.5677297910054525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,12288,0.08796622355779012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,51200,0.18796800242529976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,16384,0.06011466847525703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,51200,0.45296091503567165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,10240,0.08282311095131768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,16384,0.14630666044023302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,12288,0.04664088951216804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,10240,0.04137688875198364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,7168,0.055140445629755654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,8192,0.032920890384250216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,12288,0.1010062230957879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,6144,0.0458133320013682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,8192,0.07300977574454413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,7168,0.03165422214402093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,5120,0.04088355435265435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,7168,0.06316977739334106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,6144,0.02676711148685879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,4096,0.03174577818976508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,6144,0.054413331879509814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,5120,0.0233262214395735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,3584,0.02603911028967963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,5120,0.0469413333468967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,65536,0.48442755805121523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,3072,0.02441155579355028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,4096,0.03906044363975525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,3584,0.018072888255119324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,2560,0.02028622229894002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,3072,0.015631111131774057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,3584,0.03513688842455546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,2048,0.016255999604860943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,3072,0.031523555517196655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,2560,0.014488000008794995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,1536,0.013083555632167392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,2048,0.013810666898886362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,2560,0.026963555150561865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,1024,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,1536,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,8192,0.06267822451061673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,768,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,1536,0.018750222192870248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,1024,0.010043555663691627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,512,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,10240,0.08566844463348389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,768,0.009413333402739631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,256,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,1024,0.014627556006113688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,768,0.012548444171746572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,128,0.004582222137186262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,512,0.009024889104896123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,256,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,512,0.010746666954623329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,64,0.0042160000238153665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,32,0.004224888980388641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,128,0.008559111091825698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,256,0.009687999884287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,128,0.009328000247478485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,4096,0.020229332976871066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,64,0.009388444324334463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,32,0.009345778160625035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,65536,0.398127105500963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,51200,0.3119893338945177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,65536,0.20315378242068818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,16384,0.10185688734054565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,51200,0.15881778134240046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,16384,0.0502800014283922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,2048,0.022862222459581163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,65536,0.5167617797851562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,10240,0.0686053302552965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,16384,0.13148177994622126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,12288,0.03945866558286879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,8192,0.05406844284799364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,10240,0.03490044342146979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,12288,0.09182933304044937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,10240,0.07900444666544597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,8192,0.029323554701275293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,6144,0.03949777616394891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,8192,0.06397600306404962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,7168,0.02676444417900509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,5120,0.03275288807021247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,6144,0.023927999867333308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,6144,0.0510426660378774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,5120,0.02032444377740224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,4096,0.02755466600259145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,5120,0.04309599929385715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,3584,0.02315644423166911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,4096,0.01789777808719211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,12288,0.08021777868270874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,4096,0.03552000059021844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,3072,0.020557334025700886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,3584,0.016189333465364244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,2560,0.01797155539194743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,3584,0.032474666833877563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,51200,0.40077866448296445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,2048,0.014903111590279473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,3072,0.015104000767072042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,1536,0.012146666646003723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,3072,0.028471999698215063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,2560,0.013912000589900546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,7168,0.047305779324637524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,1024,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,2560,0.025044444534513686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,768,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,2048,0.020727111233605277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,1024,0.009736888938479954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,1536,0.017284444636768766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,512,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,7168,0.05681688918007744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,1024,0.013759999639458127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,256,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,768,0.011725333001878528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,128,0.004214222232500712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,256,0.008307555483447181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,512,0.010059555371602377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,64,0.0038942222793896994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,256,0.009031111167536842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,32,0.0038968887594011095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,128,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,128,0.008728889127572378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,64,0.008692444198661381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,32,0.00869422240389718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,65536,0.3427857822842068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,65536,0.15522400538126627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,2048,0.012729778057999082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,51200,0.2737199995252821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,1536,0.011129777464601727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,16384,0.07890844345092773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,65536,0.4332808918423123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,12288,0.052928000688552856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,51200,0.1238595512178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,768,0.009198221895429824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,16384,0.04048711061477661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,10240,0.054085334142049156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,51200,0.38732179005940753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,512,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,16384,0.12748444080352783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,8192,0.04141511188613044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,12288,0.033259554041756526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,10240,0.028635554843478735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,12288,0.08003911044862536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,6144,0.03240799903869629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,8192,0.02408355474472046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,10240,0.06720622380574544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,8192,0.05491111013624403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,6144,0.01921688848071628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,7168,0.04939111073811849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,4096,0.022005332840813532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,5120,0.01676266723208957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,3584,0.019765333996878732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,5120,0.036898665957980685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,4096,0.03109955456521776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,3584,0.013770666387346057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,3584,0.027796443965699937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,3072,0.024586665961477492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,7168,0.038132445679770574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,2560,0.014945778581831189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,3072,0.012656889028019376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,6144,0.04313688808017307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,2560,0.011687999798191918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,2048,0.01274577776590983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,4096,0.01479022204875946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,2048,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,1536,0.010337777435779572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,2048,0.018266666266653273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,1024,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,1536,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,3072,0.017337777548366122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,768,0.006646222124497096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,1536,0.015299555328157214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,1024,0.008984888593355814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,512,0.00546488869521353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,1024,0.012306667036480375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,768,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,256,0.004232888834344016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,768,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,512,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,512,0.00908088849650489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,128,0.0038088887102074097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,5120,0.027317333552572463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,64,0.003559111307064692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,256,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,128,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,2560,0.021335111724005804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,64,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,32,0.008233778178691864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,65536,0.28662223286098903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,51200,0.21443643834855822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,65536,0.14696533150143093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,65536,0.3793671131134033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,51200,0.11480177773369683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,51200,0.3256728914048937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,12288,0.05396711164050632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,16384,0.039467556609047785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,7168,0.02199288871553209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,16384,0.09187466568417019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,10240,0.04488266507784525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,32,0.003544888976547453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,12288,0.03200799889034695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,8192,0.03429866830507914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,12288,0.07067466444439359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,10240,0.02770400047302246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,7168,0.031143999762005273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,10240,0.05979022052552965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,8192,0.022329777479171753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,6144,0.025993777645958796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,8192,0.049307554960250854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,7168,0.020638222495714825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,5120,0.021440888444582622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,7168,0.04395822352833218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,6144,0.018532445033391316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,4096,0.018400000201331247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,6144,0.03864444295565287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,5120,0.016173332929611206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,3584,0.01626133256488376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,5120,0.033556444777382746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,16384,0.07102844450208876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,3072,0.014775110615624322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,3584,0.013448889056841532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,4096,0.027830223242441814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,2560,0.01237777786122428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,3584,0.025054223007626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,3072,0.012429333395428128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,2048,0.01057866629627016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,256,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,3072,0.022114665971861944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,1536,0.008459555606047312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,2048,0.010398222340477837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,1024,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,2048,0.016307555966907077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,128,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,1536,0.009380444056457942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,768,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,1536,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,1024,0.008820444345474243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,512,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,1024,0.011391110718250275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,768,0.008270222279760573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,256,0.0038888889054457345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,768,0.009760888914267222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,512,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,256,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,512,0.008569777839713627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,64,0.003142222141226133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,256,0.007748444047239091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,32,0.0031759999692440033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,128,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,4096,0.014515555567211576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,128,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,64,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,32,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,65536,0.23647554715474448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,2560,0.011420444481902652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,65536,0.1395368840959337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,2560,0.01938488913906945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,51200,0.18714133898417154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,16384,0.06044444772932264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,65536,0.3165111011928982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,12288,0.04532533221774631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,51200,0.11071822378370498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,16384,0.03627999954753452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,10240,0.03681333197487725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,51200,0.2604622311062283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,12288,0.02943555514017741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,8192,0.029462221595976088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,12288,0.060319999853769936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,10240,0.02676444417900509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,128,0.0032151111712058387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,7168,0.025612443685531616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,10240,0.05123111274507311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,6144,0.021192888418833416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,8192,0.04231289029121399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,7168,0.019345778557989333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,5120,0.018546667363908555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,6144,0.017283555534150865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,7168,0.037626667155159846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,4096,0.015790222419632804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,6144,0.03342577815055847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,3584,0.014192889134089151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,4096,0.013814222481515674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,5120,0.028582221931881372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,3072,0.012663111090660095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,4096,0.023973332511054143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,3584,0.012848888834317526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,3584,0.02145066691769494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,2560,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,3072,0.012057777908113269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,2048,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,3072,0.01934400035275353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,2560,0.011040888726711273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,1536,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,2560,0.016964443855815463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,16384,0.07923466629452176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,2048,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,1024,0.005979555762476391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,2048,0.014783110883500842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,768,0.0052933332820733385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,1536,0.009319999979601966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,1536,0.01277955538696713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,8192,0.022612444228596155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,1024,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,512,0.004552000098758274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,256,0.003548444559176763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,768,0.00870577742656072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,512,0.007814221911960179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,5120,0.015385778413878547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,128,0.00318755561278926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,256,0.006990222467316522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,64,0.0029884444342719186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,128,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,32,0.002871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,128,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,64,0.006748444504208035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,65536,0.18820977210998535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,32,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,51200,0.14719377623664007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,65536,0.23100445005628797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,16384,0.046852442953321666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,51200,0.07675022549099393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,51200,0.18114756213294136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,768,0.008013333711359236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,12288,0.0347031123108334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,512,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,16384,0.02616888946957058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,10240,0.02696711156103346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,12288,0.021412443783548143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,256,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,16384,0.05954222546683418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,8192,0.02292088833120134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,12288,0.045901334948009916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,10240,0.019360888335439894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,10240,0.039119998613993325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,8192,0.01536266671286689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,6144,0.017441777719391715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,7168,0.014155555102560254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,8192,0.0329093337059021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,5120,0.015258666541841296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,7168,0.029275556405385334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,4096,0.012699555191728802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,6144,0.01313866674900055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,5120,0.011735111474990845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,6144,0.025943999489148457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,3584,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,5120,0.022103110949198406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,4096,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,65536,0.09347999758190578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,3072,0.010265777508417765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,2560,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,4096,0.018719999326599967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,3072,0.009914666414260864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,3584,0.017210667332013447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,2560,0.009306666751702627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,2048,0.007892444729804993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,1024,0.010093332992659675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,3072,0.01567911108334859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,1536,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,2560,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,1024,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,2048,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,1536,0.008244444098737504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,2048,0.011932444241311816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,1536,0.010354667074150508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,1024,0.007687999970383114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,512,0.003899555653333664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,1024,0.008550222549173567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,768,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,256,0.003216888962520493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,768,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,512,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,128,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,512,0.006896889044178857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,256,0.006657777975002925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,64,0.0026453334010309647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,256,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,32,0.002826666666401757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,128,0.006648889018429651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,128,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,3584,0.010238221950001186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,64,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,65536,0.12399821811252171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,32,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,51200,0.10519288645850287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,65536,0.09273155530293782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,16384,0.03317600157525804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,65536,0.22234400113423666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,51200,0.07481422026952107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,12288,0.024290665984153748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,768,0.0047600000268883174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,51200,0.17462489340040419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,7168,0.020038222273190815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,10240,0.020942222740915086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,16384,0.05772622426350912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,8192,0.0174026671383116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,12288,0.02093155516518487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,7168,0.015771556231710646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,12288,0.044886221488316856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,10240,0.01904800037542979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,8192,0.015294222368134392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,6144,0.014144889182514615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,10240,0.03798933492766486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,8192,0.03173689047495524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,7168,0.014152000347773233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,5120,0.012311999996503195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,4096,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,7168,0.028400000598695543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,6144,0.013136000268989138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,3584,0.00926399976015091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,5120,0.011768889096048145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,6144,0.02480977773666382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,3072,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,4096,0.01071200023094813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,5120,0.021290666527218286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,2560,0.007518221934636434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,4096,0.018225777480337355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,3072,0.009655999640623728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,2048,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,3072,0.01516355574131012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,2560,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,1536,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,2560,0.013468444347381592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,2048,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,1024,0.004578666554556953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,2048,0.011580444044537015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,768,0.004168888760937585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,1024,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,1536,0.010094222095277574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,512,0.003528000166018804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,1024,0.008469333251317343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,768,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,256,0.002838222309947014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,512,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,512,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,128,0.002871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,256,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,256,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,64,0.0025351111673646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,128,0.006775110960006714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,3584,0.010273777776294285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,32,0.0028106667515304354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,128,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,3584,0.016703110602166917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,65536,0.09815555810928345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,64,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,51200,0.07896266380945842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,32,0.006044444524579578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,65536,0.0923688875304328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,16384,0.02937955657641093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,16384,0.025237333443429735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,12288,0.0258942229880227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,65536,0.18337599436442056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,51200,0.07467022207048205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,16384,0.024322667055659827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,51200,0.14444888962639704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,10240,0.0225608895222346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,16384,0.048489779233932495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,12288,0.01994044416480594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,768,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,8192,0.01851733360025618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,12288,0.037760890192455716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,7168,0.01313066648112403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,10240,0.018598222070270114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,10240,0.03257066673702664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,6144,0.01124533348613315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,8192,0.01480977733929952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,5120,0.009810667071077559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,7168,0.013832000394662222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,8192,0.027073777384228174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,4096,0.008236444658703273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,7168,0.024167110522588093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,5120,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,6144,0.021591110361946955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,3584,0.007510222494602203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,4096,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,5120,0.018998222218619455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,3072,0.006704888823959563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,4096,0.016023111012246873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,2560,0.00619555554456181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,3584,0.010048000348938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,3072,0.009668444593747457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,2048,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,3072,0.013381332986884646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,1536,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,1536,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,2560,0.01204622205760744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,2048,0.008392888638708327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,1024,0.0035457776652442086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,2048,0.010397333237859938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,1536,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,768,0.0032177778581778207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,1536,0.009215110705958473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,1024,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,512,0.003107555624511507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,1024,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,768,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,256,0.0025457777082920074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,768,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,6144,0.012741333080662621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,512,0.006903999795516332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,128,0.002464888824356927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,64,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,512,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,256,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,32,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,128,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,256,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,65536,0.0510062211089664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,128,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,51200,0.04160799913936191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,64,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,3584,0.01477866702609592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,65536,0.09212088584899902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,16384,0.01663644446267022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,12288,0.01423555612564087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,2560,0.008992888861232335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,65536,0.1409013271331787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,10240,0.013050666285885705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,16384,0.0236488895283805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,51200,0.11073333024978638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,12288,0.020018666982650757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,8192,0.012772444221708508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,16384,0.03816711240344577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,7168,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,12288,0.03001333276430766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,8192,0.014797333214018079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,10240,0.02602044410175747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,6144,0.009855111440022787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,7168,0.013501333693663279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,8192,0.021593777669800654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,7168,0.019674667053752475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,6144,0.012456888953844706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,4096,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,6144,0.017671111557218764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,5120,0.011423110961914062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,3584,0.006656888872385025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,5120,0.015568888849682279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,4096,0.01033866653839747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,3072,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,4096,0.013482666677898832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,32,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,2560,0.0053280000057485365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,3072,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,51200,0.07396177450815837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,3072,0.011408888631396823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,2048,0.004646222210592694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,2560,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,1536,0.003967111309369405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,2048,0.008469333251317343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,10240,0.018277333842383493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,1024,0.0034871109657817413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,1536,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,1536,0.00831288927131229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,5120,0.008657777474986183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,768,0.003108444520168834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,1024,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,1024,0.007620444728268518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,512,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,768,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,768,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,256,0.002504000026318762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,3584,0.009992000129487779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,512,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,512,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,128,0.002159111201763153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,3584,0.012448888685968188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,64,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,256,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,128,0.006647999915811751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,32,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,256,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,65536,0.04198844565285576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,128,0.005056888692908817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,51200,0.0337759984864129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,64,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,2560,0.01033333357837465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,16384,0.010551999840471478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,32,0.005031110925806893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,2048,0.008967999782827165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,65536,0.09190577930874294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,12288,0.009249777429633671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,51200,0.07358933157391019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,16384,0.023381332556406658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,10240,0.010784889260927836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,51200,0.11039200094011094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,12288,0.019742222295867074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,16384,0.0377146667904324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,12288,0.029265777932273015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,10240,0.01826844447188907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,7168,0.009232000344329411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,8192,0.014612444572978549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,10240,0.025225778420766194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,6144,0.008856000171767341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,8192,0.021314667330847845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,7168,0.013623110949993134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,5120,0.008463111188676622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,7168,0.019308444526460435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,6144,0.012674666941165924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,4096,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,6144,0.01756444407833947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,5120,0.011363555159833698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,3584,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,5120,0.015271110667122735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,4096,0.01036977767944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,3072,0.005648889061477449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,3584,0.00979288915793101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,4096,0.013245333400037555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,2560,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,3584,0.012312889099121094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,2048,0.004573333180612988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,3072,0.011384888655609555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,2560,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,1536,0.0038853333228164246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,65536,0.14085422621832952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,2560,0.010296000374688042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,1024,0.003176888864901331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,2048,0.008610666626029545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,8192,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,1536,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,1024,0.007542221910423703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,1536,0.008058666355080074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,512,0.0025928889711697898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,1024,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,768,0.007003555695215861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,256,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,768,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,512,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,128,0.0021333334346612296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,256,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,64,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,256,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,128,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,32,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,128,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,65536,0.03705155518319871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,51200,0.029904001288943823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,64,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,16384,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,32,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,3072,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,12288,0.009883555273214975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,10240,0.008867555194430882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,65536,0.14090932740105525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,8192,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,51200,0.11017688777711655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,16384,0.03729155659675598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,12288,0.029100444581773546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,7168,0.008567999634477828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,2048,0.009012444151772393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,6144,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,10240,0.025088888075616624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,5120,0.007355555891990662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,4096,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,8192,0.021337777376174927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,3584,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,7168,0.01925244430700938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,3072,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,5120,0.015409777561823526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,2560,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,512,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,4096,0.013437333206335703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,2048,0.004546666724814309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,1536,0.0038942222793896994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,3072,0.011366222467687396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,1024,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,2560,0.010051555931568146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,2048,0.00903733323017756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,768,0.0028524444335036804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,512,0.002732444347606765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,1536,0.008059555457697975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,256,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,1024,0.0069919998447100324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,128,0.0021199999998013177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,768,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,512,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,64,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,32,0.0018639999131361644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,256,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,768,0.0028568889117903183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,65536,0.039135111702813044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,128,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,51200,0.03042311138576931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,16384,0.009479111267460717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,64,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,32,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,12288,0.011680000358157687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,10240,0.010592888626787396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,6144,0.017468444175190397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,65536,0.1399573352601793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,8192,0.00961866643693712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,7168,0.008832889298597971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,51200,0.10977155632442898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,16384,0.03734933336575826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,6144,0.012615111139085559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,3584,0.012275555895434486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,5120,0.011306666665607028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,10240,0.025177778469191656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,4096,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,3584,0.007247111035717859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,8192,0.021254221598307293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,3072,0.006504888749784893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,7168,0.019325332509146798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,6144,0.01755555636352963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,2560,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,5120,0.015406222807036506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,2048,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,4096,0.013386666774749756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,1536,0.0036195553839206696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,3584,0.012413333687517377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,1024,0.0031537777847713898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,768,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,3072,0.011370666325092316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,512,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,2560,0.010103999740547603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,256,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,2048,0.009155555731720394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,128,0.0018666666001081467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,1536,0.008031999899281396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,1024,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,64,0.0018213333355055915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,32,0.0018275556051068837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,512,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,256,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,128,0.005068444543414646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,64,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,32,0.00499555551343494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,12288,0.029167999823888142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,768,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,16384,1.9994612799750433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,16384,0.611386670006646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,16384,2.22090933057997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,51200,2.1614952087402344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,12288,1.5017307069566515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,16384,1.8285022311740453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,12288,1.5683920118543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,12288,0.45720709694756406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,10240,1.1430959701538086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,10240,0.38788265652126735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,12288,1.3547635608249242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,10240,1.2596569061279297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,8192,0.3266773223876953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,51200,6.341093275282119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,10240,1.0892808702256944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,8192,1.0618089040120442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,7168,0.8572470876905652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,7168,0.2852568891313341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,8192,0.8435191048516167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,8192,0.9769289228651258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,7168,0.8608018027411567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,6144,0.7104098002115885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,51200,8.056445651584202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,6144,0.7644711070590549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,7168,0.7575048870510526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,5120,0.6433084275987413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,5120,0.6298142009311253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,5120,0.21417689323425293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,6144,0.6372115347120497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,4096,0.5628817876180013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,5120,0.5399920145670573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,4096,0.18073511123657227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,6144,0.2529582182566325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,4096,0.48094489839341903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,3584,0.4456613328721788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,3584,0.16047110822465685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,4096,0.4432071050008138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,3584,0.4789582358466254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,3072,0.38794223467508954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,3584,0.3895724349551731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,3072,0.14198222425248888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,2560,0.27582934167650014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,2560,0.12435732947455512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,3072,0.3443733321295844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,2560,0.31170132425096303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,2048,0.23851911226908365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,1536,0.17897866831885445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,2048,0.10493155320485432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,2560,0.2877226670583089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,3072,0.3199342091878255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,2048,0.2529919942220052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,1024,0.1321208874384562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,1536,0.20506666766272652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,1536,0.08384888701968723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,2048,0.24154755804273817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,51200,6.008856031629775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,1024,0.06774222188525729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,512,0.0753564437230428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,1024,0.15216622087690565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,1536,0.18794577651553682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,768,0.07017688618765937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,768,0.12310400274064805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,256,0.06155288881725735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,1024,0.1417671177122328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,512,0.05325422353214688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,768,0.11744711134168838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,128,0.043528887960645885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,512,0.10786577728059556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,256,0.04890399840142992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,256,0.0946719977590773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,64,0.03977333174811469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,128,0.047800001170900136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,512,0.09794844521416558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,128,0.08864177597893609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,128,0.05870933002895779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,256,0.07731911208894518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,32,0.0888177752494812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,64,0.0878524449136522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,768,0.09620977772606744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,32,0.03894666830698649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,65536,2.112043592664931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,51200,1.6451404359605577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,51200,4.986479017469618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,16384,1.9181342654758031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,16384,1.7165555953979492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,65536,6.563326093885634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,51200,6.154784732394749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,16384,0.5075822406344944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,12288,0.36934577094184023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,12288,1.2928684022691515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,12288,1.1627368927001953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,65536,8.60093519422743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,16384,1.326284408569336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,10240,1.1765058305528429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,12288,1.0504035949707031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,10240,0.964646233452691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,10240,0.3095422320895725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,51200,4.679793887668185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,8192,0.25588801172044545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,8192,0.7666897773742676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,8192,0.8453164630466037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,7168,0.22192888789706758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,7168,0.8025760120815701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,10240,0.8704862064785428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,7168,0.7270284228854709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,8192,0.6715155707465278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,6144,0.6359448962741429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,7168,0.5924613210890028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,6144,0.19749155309465197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,6144,0.6146737734476725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,5120,0.4795386526319716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,5120,0.16750400596194795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,4096,0.40146578682793516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,5120,0.5009119775560167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,4096,0.14033955997890896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,4096,0.38340266545613605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,3584,0.34792621930440265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,6144,0.5116622183057997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,5120,0.452439096238878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,65536,6.112692515055339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,3584,0.33325245645311147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,4096,0.3497813277774387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,3584,0.12565333313412136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,3072,0.29206665356953937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,2560,0.21325600147247314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,3072,0.3108328978220622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,3072,0.12562666998969182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,3584,0.3107084433237712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,2560,0.26502754953172475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,2048,0.19452444712320963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,3072,0.2673039966159397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,1536,0.14276711146036783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,2560,0.2287022272745768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,2560,0.0976453357272678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,2048,0.23144088851081002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,2048,0.0821582211388482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,1024,0.11082755857043797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,1536,0.06619999806086223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,1536,0.1593831115298801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,2048,0.19101511107550728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,768,0.07740444607204862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,1024,0.053263111246956714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,1536,0.15282666683197021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,768,0.04758222235573662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,768,0.09672800037595962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,1024,0.11559466520945232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,256,0.0412124428484175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,768,0.09513955646091038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,512,0.04322133461634318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,512,0.08354844649632771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,128,0.03325066632694668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,256,0.039870222409566246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,256,0.07460355758666992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,512,0.07872888776991102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,64,0.030736890104081895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,128,0.07062488794326782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,32,0.02948088778389825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,256,0.06248799959818522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,128,0.04801511102252536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,64,0.07007377677493624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,1024,0.12188177638583714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,32,0.0707226660516527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,512,0.06646755668852064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,128,0.03832799858517117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,65536,0.6753786404927572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,51200,1.5936898125542536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,65536,2.059962590535482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,16384,0.6328373485141331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,51200,0.5491351021660699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,51200,1.8239786359998915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,16384,0.17711110909779867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,12288,0.42731910281711155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,16384,0.5375608868069118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,65536,2.7400889926486545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,12288,0.14145421981811523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,12288,0.3820879989200168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,10240,0.3268088764614529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,16384,0.4499048656887478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,10240,0.10929422246085273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,8192,0.2651253276401096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,51200,1.5532986323038738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,10240,0.3373200098673503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,12288,0.33748711480034727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,8192,0.2492542266845703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,8192,0.09882666667302449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,7168,0.244832886589898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,10240,0.2914995617336697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,7168,0.07920266522301568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,65536,2.0347796546088324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,7168,0.2244124412536621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,8192,0.231894228193495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,7168,0.20203555954827201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,5120,0.17567377620273164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,6144,0.20668978161282012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,4096,0.14287111494276258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,5120,0.0590533349249098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,6144,0.17620621787177193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,5120,0.16057066122690836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,3584,0.11728266874949138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,4096,0.049696889188554555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,4096,0.1304488844341702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,5120,0.15045155419243708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,3072,0.09986222452587551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,3584,0.11282311545477973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,3584,0.04454844527774387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,4096,0.13019733958774143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,2560,0.0894177754720052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,3072,0.039903998374938965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,3584,0.11092266771528457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,3072,0.09903822342554729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,6144,0.22027999824947783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,2560,0.035608000225490995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,2560,0.08703733152813381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,3072,0.09799910916222467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,1536,0.04683822062280443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,2048,0.030264887544843886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,2560,0.08569511440065171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,2048,0.07058755556742351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,1024,0.03450399968359206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,1536,0.0243520008193122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,1536,0.05602044529385037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,2048,0.07175555494096544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,768,0.025340444511837427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,1024,0.042931556701660156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,1024,0.02058844433890449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,1536,0.060306668281555176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,512,0.020544888244734872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,768,0.03485866718822055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,768,0.018419555491871305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,512,0.017082666357358296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,768,0.04048711061477661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,256,0.014666666587193808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,512,0.028588443994522095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,2048,0.07059999969270495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,256,0.01566133399804433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,512,0.033775111039479576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,128,0.014339556296666464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,256,0.025753777888086107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,6144,0.06954222255282931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,64,0.0120000003112687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,32,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,128,0.01553066737122006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,256,0.02752888864941067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,128,0.02464977734618717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,64,0.024466666910383437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,128,0.021418665846188862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,32,0.02456088860829671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,1024,0.047919111119376294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,65536,0.51382933722602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,65536,1.3306728998819988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,51200,1.095671123928494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,16384,0.36411910586886936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,51200,0.398693323135376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,16384,0.4606311056349013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,65536,2.1364453633626304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,16384,0.13487911224365234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,12288,0.10634489191903009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,12288,0.2888186772664388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,16384,0.3460871113671197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,51200,1.125791973537869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,10240,0.25406400362650555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,12288,0.26133423381381565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,65536,1.482217788696289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,10240,0.09067200289832221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,8192,0.20362755987379286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,51200,1.424438264634874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,8192,0.06862044334411621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,7168,0.17675999800364176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,10240,0.21728266610039604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,8192,0.20173954963684082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,12288,0.23453866110907662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,6144,0.1369475523630778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,7168,0.16752089394463432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,8192,0.18577778339385986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,7168,0.06616799698935615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,5120,0.11089333560731675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,6144,0.05396533343527052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,6144,0.14867466025882298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,7168,0.1554622252782186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,4096,0.09499377674526638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,5120,0.04616711205906338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,5120,0.12586843967437744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,6144,0.13628621896107992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,4096,0.03867022196451823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,3584,0.08712000317043728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,4096,0.0991777777671814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,5120,0.12355111704932319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,3072,0.07065244515736897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,3584,0.03403466608789232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,10240,0.24037334654066297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,4096,0.09796888960732354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,3072,0.030200001266267564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,3072,0.07674044370651245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,3584,0.08832710981369019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,2560,0.027570666538344488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,2560,0.06620089213053386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,2560,0.06842044326994154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,2048,0.023844444089465674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,2048,0.055493334929148354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,1536,0.036551998721228704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,2048,0.058014220661587186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,1536,0.02022133270899455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,1024,0.02586577832698822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,1536,0.04385777645640903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,1536,0.04861155483457777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,1024,0.017163554827372234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,768,0.01907999979125129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,1024,0.03429422113630507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,3584,0.08732710944281684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,2560,0.06361066632800631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,512,0.01498044364982181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,1024,0.03847288754251268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,768,0.015253333581818474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,256,0.010965333216720156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,2048,0.046661333905326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,512,0.014018666413095264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,3072,0.07911199993557401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,512,0.022831110490692988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,128,0.010250666903124915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,256,0.013008000122176277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,256,0.02032533288002014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,64,0.008782222039169734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,512,0.028040889236662123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,32,0.008938666847017076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,128,0.013073777986897362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,128,0.01958933307064904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,256,0.022112889422310725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,64,0.0196435550848643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,128,0.01763288842307197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,32,0.01960088809331258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,768,0.027806222438812256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,65536,0.4551555315653483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,65536,1.2202879587809246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,51200,0.9052498075697156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,16384,0.3282817734612359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,65536,1.747579574584961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,51200,0.35752354727851016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,51200,1.2351030773586698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,768,0.032484445306989886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,16384,0.11309688621097141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,16384,0.33549777666727704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,12288,0.2600604428185357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,65536,1.2767723931206596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,16384,0.30425066418117946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,12288,0.08525511291291978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,10240,0.2289057837592231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,12288,0.24043644799126518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,8192,0.1761066648695204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,10240,0.07290399736828275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,12288,0.22889243231879342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,51200,0.9957830641004773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,8192,0.17855377991994223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,7168,0.16398844454023573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,8192,0.05845955345365736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,10240,0.1928346686893039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,6144,0.13560799757639566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,7168,0.052595555782318115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,8192,0.15528355704413518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,7168,0.1439244482252333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,5120,0.11662044790056016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,6144,0.04675999946064419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,6144,0.12261688709259033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,4096,0.0918204453256395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,7168,0.14914045068952772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,5120,0.041464000940322876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,5120,0.10435377889209324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,6144,0.12043378088209365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,3584,0.08068888717227511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,4096,0.0336773329310947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,4096,0.08376622200012207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,5120,0.1037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,3072,0.06830666462580363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,3584,0.030476444297366675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,2560,0.05911111169391208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,3584,0.07458933194478352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,4096,0.08723466926150852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,3072,0.027152889304690894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,3072,0.06521066692140368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,10240,0.23698311381869844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,2048,0.04778844449255201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,3584,0.07803377840254042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,2560,0.024855110380384658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,1536,0.034112887250052557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,2048,0.02127911150455475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,3072,0.07000178098678589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,1024,0.0251555558707979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,2048,0.046893331739637584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,1536,0.017647999856207106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,1536,0.03817955652872721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,2560,0.06163110997941759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,768,0.01977688901954227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,2048,0.05259022116661072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,1024,0.029517332712809246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,1536,0.043154666821161904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,512,0.01758311192194621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,768,0.013853333062595792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,768,0.02436622149414486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,1024,0.03445777628156874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,256,0.013316444224781461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,512,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,512,0.01995111174053616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,768,0.029746668206320867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,128,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,256,0.012087110843923358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,256,0.017536888519922893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,64,0.01035288886891471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,128,0.011752000285519494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,256,0.020167110694779288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,128,0.01720711092154185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,32,0.01092355532778634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,2560,0.05676888757281833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,128,0.0166631109184689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,64,0.01719111038578881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,32,0.017112000121010672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,1024,0.014877332581414117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,65536,0.37105867597791886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,65536,1.0143377516004775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,51200,0.8560809029473199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,512,0.025610667135980394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,51200,0.31107378005981445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,51200,0.9105200237698026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,65536,1.3468507130940754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,16384,0.33052089479234487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,12288,0.181329780154758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,16384,0.10496888558069865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,16384,0.2582062350379096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,12288,0.22398045327928331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,12288,0.07601600223117404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,10240,0.17212443881564668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,51200,0.8136622111002604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,65536,1.0536923938327365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,12288,0.1978764401541816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,8192,0.1432222260369195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,10240,0.16968622472551134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,7168,0.11431822511884902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,10240,0.16737866401672363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,8192,0.05158311128616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,16384,0.2751164436340332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,6144,0.09278933207194011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,7168,0.04646133383115133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,8192,0.13575643963283962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,7168,0.11867466237809922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,5120,0.08065599865383573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,6144,0.038146668010287814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,7168,0.12077066633436416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,6144,0.10090221961339314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,4096,0.0737982193628947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,5120,0.033783111307356097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,6144,0.10739288727442424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,10240,0.06059910853703817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,4096,0.028885334730148315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,3584,0.059768888685438365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,5120,0.08994222349590725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,8192,0.14828977319929335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,3584,0.02604888876279195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,4096,0.07642844650480482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,3072,0.0520648890071445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,3584,0.06204355425304837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,2560,0.04383644461631775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,3072,0.023801777097913954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,3584,0.06943821907043457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,3072,0.054218666421042554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,2048,0.03419288992881775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,2560,0.021594666772418555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,3072,0.061055110560523145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,1536,0.026141333911154006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,2048,0.01918577816751268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,2048,0.03915111223856608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,2560,0.05429600013626946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,5120,0.08628800180223252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,1024,0.01867466668287913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,2048,0.04615644282764859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,1536,0.01648533344268799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,1536,0.03225244416130914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,768,0.015020444989204407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,4096,0.0696444445186191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,1024,0.013925333817799887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,1024,0.02494666642612881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,512,0.011845333708657159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,1536,0.03774133324623108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,768,0.012740444805887012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,768,0.020868443780475195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,1024,0.030731555488374498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,512,0.011698666546079846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,512,0.016922665966881644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,768,0.02646133303642273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,128,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,256,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,256,0.015266666809717814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,512,0.022153778208626643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,64,0.00869511150651508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,32,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,128,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,128,0.014826666977670459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,256,0.018332444959216647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,2560,0.047129776742723256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,64,0.0148044443792767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,128,0.015258666541841296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,32,0.014780445231331719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,65536,0.329128000471327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,256,0.011231111155615913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,65536,0.9451395670572916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,16384,0.2538248962826199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,51200,0.25516266292995876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,51200,0.8082097901238335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,65536,1.0163582695855033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,12288,0.18890399403042266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,16384,0.08712177806430393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,16384,0.2511928876241048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,51200,0.6975902451409234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,12288,0.18533778190612793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,12288,0.06400266620847914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,16384,0.22511111365424263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,65536,0.9012364281548394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,8192,0.12997066974639893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,10240,0.05431110991372002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,12288,0.17875911129845512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,8192,0.04355822337998284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,7168,0.10940621958838569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,8192,0.11824266115824382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,10240,0.14570666684044734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,6144,0.09415111276838516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,7168,0.03878488805558946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,8192,0.12008710702260335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,7168,0.10429688957002427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,6144,0.09000355667538112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,7168,0.10776711172527736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,6144,0.03398666779200236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,5120,0.07892000012927584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,10240,0.16041599379645452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,6144,0.09454933140012954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,4096,0.06671910815768771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,5120,0.029715554581748113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,5120,0.07724089092678495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,3584,0.05461777912245857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,4096,0.025455999705526564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,10240,0.15711822774675158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,5120,0.08017510837978788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,4096,0.06325333648257785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,3072,0.04883111185497708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,51200,0.6927848921881782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,3584,0.023579556081030104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,4096,0.06659111049440172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,3584,0.056778669357299805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,2560,0.04098844528198242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,3072,0.021535999245113794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,3072,0.04964444372389051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,2560,0.019505777292781405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,3584,0.060439109802246094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,2560,0.04294755392604404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,1536,0.0258568889564938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,2048,0.01721955504682329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,2560,0.04833244283994039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,2048,0.036185777849621244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,1024,0.01886666648917728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,1536,0.014530667000346713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,2048,0.04128977656364441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,1024,0.012673777838548025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,1536,0.03421777817938063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,1024,0.02274488906065623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,1024,0.027832888894610937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,768,0.018811555372344125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,512,0.010631110933091907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,768,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,768,0.02368444369898902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,2048,0.03365600109100342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,256,0.008242666721343994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,512,0.01534400052494473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,3072,0.054654220740000405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,128,0.007187555233637492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,256,0.014057776994175382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,256,0.010363555616802638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,1536,0.029345777299669053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,64,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,128,0.01038933296998342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,32,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,768,0.016422222057978313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,128,0.014470222923490735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,64,0.01348888874053955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,32,0.013454222016864352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,512,0.010788444843557147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,65536,0.27316000726487905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,65536,0.7084933386908637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,512,0.0203155560625924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,256,0.01740711099571652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,51200,0.47017865710788304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,51200,0.21588355965084502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,16384,0.1881528827879164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,16384,0.06927733288870917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,65536,0.7899635632832845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,16384,0.20603466033935547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,12288,0.14375821749369302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,16384,0.18900711006588408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,12288,0.053583112027910024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,12288,0.17235555913713244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,10240,0.12020266056060791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,51200,0.5850488874647353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,8192,0.09466489156087239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,12288,0.14586044682396784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,10240,0.04548888736300998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,128,0.013438222308953604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,10240,0.1270995537439982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,65536,0.9486249287923177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,7168,0.0893297791481018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,8192,0.03638933433426751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,8192,0.10380444261762832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,6144,0.07598577605353461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,7168,0.03275110986497667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,7168,0.09111733569039239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,8192,0.10346488820181952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,5120,0.06382044156392415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,6144,0.029273778200149536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,6144,0.08001422219806247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,7168,0.09153955512576634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,4096,0.04745777779155307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,5120,0.025973333252800837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,6144,0.08113511403401692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,5120,0.06810222069422404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,4096,0.022525333695941504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,5120,0.0700026684337192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,3584,0.04976088802019755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,3584,0.020640888147883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,3072,0.03677511215209961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,3584,0.05276266733805338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,10240,0.12665955225626627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,3072,0.04373244444529215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,2560,0.030065778228971694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,3072,0.01903733279969957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,51200,0.6932951079474555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,3072,0.04790488878885905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,2048,0.025363556212849084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,2560,0.017128000656763714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,2560,0.03812177644835578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,3584,0.04191911220550537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,1536,0.01958577831586202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,2560,0.04127733243836297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,4096,0.05586222145292494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,1536,0.013457777599493662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,2048,0.03586577706866794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,4096,0.05891288651360405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,1024,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,1536,0.026514665948020086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,768,0.011385777758227454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,1024,0.011418666276666852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,1024,0.020125332805845473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,768,0.016897777716318767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,512,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,768,0.01072799993885888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,768,0.02125866711139679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,512,0.013958222336239286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,256,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,512,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,2048,0.03218044506178962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,512,0.018518222702874076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,128,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,256,0.01278222186697854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,256,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,64,0.005374222166008419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,128,0.012418666647540199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,32,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,256,0.016571554872724745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,1536,0.029998222986857098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,64,0.0123831108212471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,32,0.012198222180207571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,1024,0.02458222210407257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,65536,0.5557031101650661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,65536,0.25506843460930717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,2048,0.015119999647140503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,51200,0.4961084259880914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,65536,0.7737181981404623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,16384,0.1640862226486206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,128,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,51200,0.203947557343377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,12288,0.1273697747124566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,128,0.013427555561065674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,16384,0.06311911344528198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,16384,0.18300622039371064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,65536,0.7516213523017036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,10240,0.10923466417524551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,12288,0.050388445456822716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,16384,0.19622577561272514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,12288,0.1297502252790663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,51200,0.5677422417534722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,10240,0.04347199863857693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,10240,0.1188160048590766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,12288,0.14222578207651773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,8192,0.03529422150717841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,10240,0.12082489331563313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,7168,0.031601776679356895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,8192,0.10690755314297146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,7168,0.0790355536672804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,6144,0.06882399982876249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,7168,0.08971199724409316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,51200,0.6169404453701443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,6144,0.028190222051408555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,5120,0.05392000079154968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,6144,0.06833333439297147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,4096,0.043450666798485645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,5120,0.025070221887694463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,6144,0.07756800121731229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,5120,0.05935200055440267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,8192,0.088009774684906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,3584,0.03708622190687392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,4096,0.04767733481195238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,7168,0.0718640022807651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,3072,0.033424887392256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,3584,0.01980533368057675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,4096,0.05758844481574165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,3584,0.042865776353412204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,2560,0.028519110547171697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,3584,0.05206222335497538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,3072,0.037745777103636004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,2560,0.033139556646347046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,3072,0.04555288950602213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,2048,0.023784889115227595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,2560,0.040807998842663236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,2048,0.027767111857732136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,1536,0.01871200071440803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,2048,0.01482399966981676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,2048,0.03496266735924615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,5120,0.06664088699552748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,1024,0.014646222194035849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,1536,0.013078221844302284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,8192,0.08927910857730442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,1024,0.01755822201569875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,1536,0.029431111282772485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,3072,0.018223999275101554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,1024,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,1024,0.02349511120054457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,2560,0.016473778420024447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,768,0.015251555376582675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,768,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,768,0.020820443828900654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,512,0.012444444828563266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,256,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,512,0.009762666291660732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,4096,0.021653332644038733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,512,0.0181377778450648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,128,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,1536,0.0224906669722663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,256,0.011542222566074796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,256,0.00925688859489229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,64,0.0057297779454125305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,128,0.009390222529570261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,128,0.011090666883521609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,32,0.0058711109062035876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,256,0.015815999772813585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,768,0.011896889242861005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,64,0.011058666639857821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,128,0.012423999607563019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,32,0.011050666371981302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,512,0.008944888909657797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,65536,0.53638580110338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,65536,0.20710489484998915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,51200,0.4041564464569092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,51200,0.1646711164050632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,16384,0.1418773333231608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,65536,0.6271386676364475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,12288,0.10721333159340753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,16384,0.0522382226255205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,51200,0.47917954126993817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,16384,0.14334489239586726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,65536,0.5853590965270996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,10240,0.09877244631449382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,12288,0.040756444136301674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,12288,0.1128328906165229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,51200,0.4577617645263672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,16384,0.15635200341542563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,8192,0.07096088594860502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,10240,0.037444445821974016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,7168,0.057427558634016246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,12288,0.12694756189982095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,8192,0.07727377944522433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,10240,0.1054240001572503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,7168,0.027298667364650305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,7168,0.07093777921464708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,8192,0.08632444673114353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,5120,0.04526399903827243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,6144,0.02369244396686554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,7168,0.07521510786480375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,6144,0.05956355730692545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,4096,0.03689422210057577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,5120,0.021168889270888433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,5120,0.050316443045934044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,4096,0.018523555662896898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,5120,0.056800888644324414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,4096,0.04190399911668566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,3584,0.03312177790535821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,3584,0.0169804443915685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,10240,0.09224355883068508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,3584,0.03731911049948798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,6144,0.056405332353379994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,3584,0.043621334764692515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,2560,0.023762666516833838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,3072,0.033088889386918806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,3072,0.015632000234391954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,3072,0.038943999343448214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,2560,0.01427288850148519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,2048,0.01994311147265964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,6144,0.0662319991323683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,2560,0.034330666065216064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,2048,0.013603555659453074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,2048,0.02438933319515652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,1536,0.01351999988158544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,2048,0.030174222257402208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,1024,0.010496888723638324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,1536,0.011399110986126794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,4096,0.0483457777235243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,1536,0.019955555597941082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,8192,0.02900800108909607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,768,0.008620444271299574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,1024,0.010025777750545079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,1536,0.02440533373090956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,1024,0.015570667054918079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,768,0.009456000394291347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,768,0.01312888910373052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,1024,0.02008800043000115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,768,0.018214222457673814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,256,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,512,0.008929777476522658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,2560,0.029041777054468792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,128,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,256,0.008610666626029545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,512,0.016902221573723685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,256,0.010079111489984725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,128,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,128,0.00847111145655314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,64,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,32,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,128,0.011775111158688864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,3072,0.02881511052449544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,64,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,32,0.009708444277445475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,512,0.0068862222962909276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,65536,0.4742550849914551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,65536,0.1961137718624539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,51200,0.3787208928002252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,65536,0.571064896053738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,16384,0.120615111456977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,51200,0.1561386717690362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,256,0.013914666242069669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,51200,0.4483493169148763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,65536,0.5761093563503689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,12288,0.09089866611692642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,16384,0.05025777882999844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,16384,0.13828889528910318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,51200,0.45663197835286456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,10240,0.08492533365885417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,12288,0.03982933362325033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,16384,0.1562239991294013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,8192,0.06765777534908719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,10240,0.03459555572933621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,12288,0.11745599905649821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,10240,0.08500177992714776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,7168,0.060030221939086914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,8192,0.027906666199366253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,10240,0.09901422262191772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,8192,0.06892177793714735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,6144,0.05042755603790283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,7168,0.025813332862324182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,8192,0.08145511150360107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,7168,0.061736000908745654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,5120,0.04223377837075127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,6144,0.023373333944214716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,6144,0.05315733287069532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,7168,0.07271022266811795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,4096,0.0340302222304874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,5120,0.02085777786042955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,5120,0.046388443973329335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,6144,0.06514666477839152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,3584,0.030952887402640447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,12288,0.10008888774447972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,4096,0.017966222431924608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,3072,0.02648977769745721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,4096,0.03814311159981622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,3584,0.016551110479566786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,3584,0.034966223769717746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,2560,0.022476444641749065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,4096,0.04669777883423699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,3072,0.01516533394654592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,3584,0.0429111123085022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,3072,0.030983110268910725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,2048,0.019021333919631112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,2560,0.013990221752060784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,2560,0.026820444398456152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,3072,0.03840444485346476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,1536,0.01482311056719886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,2048,0.012490666574902005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,2048,0.02269511090384589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,1024,0.010791999598344168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,1536,0.01854222185081906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,1536,0.011295110815101199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,2048,0.02899200055334303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,768,0.008233778178691864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,1536,0.024081778195169237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,768,0.012451555165979596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,5120,0.055481778250800244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,1024,0.02013066742155287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,512,0.006215111249023014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,768,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,512,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,256,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,768,0.018010665973027546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,512,0.008818666968080733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,512,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,128,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,256,0.008441777692900764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,256,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,512,0.0155102229780621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,64,0.004179555508825514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,2560,0.033951110310024686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,32,0.004268444246715969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,128,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,128,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,64,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,128,0.011401777466138205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,65536,0.39585423469543457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,65536,0.15952089097764757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,65536,0.4832906723022461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,51200,0.3252764542897542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,51200,0.13188710477617052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,65536,0.4463466538323297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,51200,0.3844630983140733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,256,0.013728000223636627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,16384,0.10251200199127197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,16384,0.04318311148219638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,1024,0.014496889379289416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,16384,0.11528444290161133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,12288,0.0790133343802558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,16384,0.12114489078521729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,12288,0.032088889016045466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,10240,0.06107378005981445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,12288,0.08756622340944077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,10240,0.028177777926127117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,12288,0.09172355466418797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,10240,0.07555555635028414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,8192,0.04833333359824287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,8192,0.022848000129063923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,10240,0.07913244432873197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,7168,0.04248977700869242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,8192,0.060247997442881264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,8192,0.06558044751485188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,51200,0.35282489988538956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,7168,0.021032000581423443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,6144,0.03631022241380479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,32,0.009200000100665623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,5120,0.030797332525253296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,6144,0.01889600025282966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,7168,0.05884266561932034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,6144,0.046929776668548584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,4096,0.02497688929239909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,5120,0.04054044352637397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,6144,0.051854223012924194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,3584,0.022026666336589392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,4096,0.01482311056719886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,4096,0.0337137778600057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,5120,0.045112000571356885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,3072,0.0191600008143319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,3584,0.013805333111021252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,4096,0.038887110021379255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,3584,0.030571556753582422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,2560,0.016526222229003906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,3072,0.01278133359220293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,3584,0.03497777713669671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,7168,0.054346667395697705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,1024,0.009940444595283931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,2048,0.014463999205165438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,2560,0.011694221860832639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,2560,0.023398222194777593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,3072,0.03156177865134345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,1536,0.011391999820868174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,2048,0.010496888723638324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,2560,0.028417779339684382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,2048,0.019687111179033916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,1024,0.008623999853928884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,1536,0.016495111915800307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,1536,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,5120,0.01716444392999013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,768,0.007287110719415877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,2048,0.02420444455411699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,1024,0.008980444735950893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,1024,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,512,0.006002666635645761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,768,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,768,0.011422221859296164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,256,0.004642666627963384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,1024,0.01721244388156467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,512,0.008106666306654612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,768,0.01551466683546702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,512,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,128,0.003903111235962974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,256,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,256,0.009036444127559662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,512,0.01418133411142561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,64,0.003549333247873518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,32,0.0038488888078265717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,128,0.007811555431948767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,128,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,256,0.01309511148267322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,64,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,128,0.010395555860466426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,32,0.008692444198661381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,65536,0.3068924480014377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,65536,0.1525626712375217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,1536,0.020505777663654752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,65536,0.4288497765858968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,16384,0.07263466384675768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,51200,0.33470577663845485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,51200,0.12167644500732422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,12288,0.06852622164620294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,16384,0.03951999876234267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,16384,0.10326755709118313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,3072,0.027246221899986267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,51200,0.3449520005120172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,10240,0.05366222063700358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,12288,0.03149155444569058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,12288,0.07738933298322889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,16384,0.11712800131903754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,8192,0.04449866546524895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,10240,0.028008888165156048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,10240,0.06733955277336968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,12288,0.09048266543282403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,7168,0.038402666648228966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,8192,0.022656889425383672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,10240,0.07713244358698527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,51200,0.26723021931118435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,6144,0.03330399923854404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,7168,0.04898311032189263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,7168,0.020643555455737643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,8192,0.06497599681218465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,5120,0.02805333336194356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,7168,0.058127111858791776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,6144,0.018428444862365723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,6144,0.042709334029091724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,65536,0.4361902342902289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,4096,0.02288088864750332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,5120,0.016753777861595154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,5120,0.03671911027696397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,6144,0.05112266540527344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,3584,0.020786666207843356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,4096,0.014411555396185981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,4096,0.030932443009482488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,5120,0.04461155666245354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,3072,0.018565333551830716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,3584,0.013739555246300168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,3584,0.02786133355564541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,4096,0.037575112448798284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,2560,0.01589866644806332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,3072,0.01257955531279246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,3072,0.024648000796635944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,2048,0.013721778160995908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,2560,0.02148355543613434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,2560,0.011503111157152386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,3072,0.03132355544302199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,1536,0.011100444528791638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,2560,0.028018666638268366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,2048,0.01833599971400367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,1536,0.01516533394654592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,1536,0.009726222190592024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,1024,0.008547555241319869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,2048,0.023659555448426142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,1024,0.008963555925422246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,1536,0.020473778247833252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,768,0.007615110940403408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,1024,0.012273777690198688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,768,0.008336000144481659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,1024,0.01721244388156467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,768,0.010576888918876648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,512,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,3584,0.034279998805787824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,256,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,8192,0.05398755603366428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,512,0.008036444584528605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,512,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,128,0.004179555508825514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,256,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,512,0.014140443669425117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,2048,0.010422222316265106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,256,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,64,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,32,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,128,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,128,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,256,0.012537777423858643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,64,0.007915555602974361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,32,0.008050666915045844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,128,0.010107555323176915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,65536,0.27344266573588055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,65536,0.1425644424226549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,51200,0.21722043885125053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,65536,0.37198400497436523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,16384,0.07288266552819146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,51200,0.2910906738705105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,16384,0.037666665183173284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,16384,0.08941688802507188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,12288,0.05362400081422594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,768,0.015454222758611044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,10240,0.04513955447408888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,65536,0.432020452287462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,51200,0.33904621336195206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,12288,0.03002133303218418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,16384,0.11675200197431777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,8192,0.0347946650452084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,12288,0.06909244590335421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,10240,0.0269795556863149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,7168,0.0314044455687205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,8192,0.021817778547604878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,12288,0.08980800045861138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,8192,0.048672888014051646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,10240,0.0773271123568217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,6144,0.026687999566396076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,7168,0.019911999503771465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,7168,0.043493333790037364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,8192,0.06400977902942233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,5120,0.02292266653643714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,6144,0.0177831103404363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,51200,0.11305422253078884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,6144,0.038254221280415855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,4096,0.01927288870016734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,7168,0.056983113288879395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,5120,0.03315022256639268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,5120,0.01593244406912062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,6144,0.05045688814587063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,4096,0.014125333891974555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,5120,0.04424444503254361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,3584,0.024672889047198828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,4096,0.0373911095990075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,3584,0.013105777402718862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,3584,0.033394667837354876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,10240,0.0587422251701355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,2560,0.013334222137928009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,3072,0.02200177808602651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,3072,0.01238933371173011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,3072,0.03012088934580485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,2560,0.011419555379284753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,2560,0.019101333287027147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,2048,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,2048,0.010112000008424124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,1536,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,2560,0.027073777384228174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,2048,0.016537777251667447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,3584,0.016903110676341586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,1024,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,2048,0.023197333017985027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,1536,0.009372444616423713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,4096,0.02761777738730113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,1536,0.013795555465751223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,768,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,1024,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,3072,0.015125332607163323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,512,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,768,0.008344000412358178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,1024,0.01683644453684489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,768,0.015470221638679504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,256,0.003882666842805015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,512,0.008492444124486711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,512,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,256,0.0075839997993575195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,512,0.013827555709415011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,128,0.003565333369705412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,256,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,256,0.0127688886390792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,64,0.003265777809752358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,128,0.007376888559924231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,32,0.0034871109657817413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,128,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,65536,0.17640444967481825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,64,0.007330666813585494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,1024,0.011380444798204633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,32,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,1536,0.020012444920010038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,768,0.009704888694816166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,51200,0.15909777747260198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,65536,0.09718666474024455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,16384,0.054261333412594266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,65536,0.2900933424631755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,51200,0.07721866501702203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,12288,0.04089777668317159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,51200,0.22807910707261828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,16384,0.02639111048645443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,16384,0.07152266634835137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,10240,0.032111111614439226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,65536,0.2971155643463135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,51200,0.23257154888576934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,12288,0.021528000632921856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,16384,0.08226488696204291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,8192,0.027309333284695942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,12288,0.0548622210820516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,128,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,10240,0.01960533360640208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,7168,0.023842665884229872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,12288,0.06423999865849812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,8192,0.015825778245925903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,10240,0.04665155543221367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,6144,0.019875556230545044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,8192,0.038853334056006536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,7168,0.014492443866199918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,10240,0.05504622062047323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,7168,0.034781333472993635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,5120,0.017092444830470614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,8192,0.04593955477078756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,6144,0.013361777696344586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,6144,0.03084533413251241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,4096,0.013674666484196981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,7168,0.04210933380656772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,5120,0.012104888757069906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,3584,0.012372444073359171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,4096,0.011032889286677042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,6144,0.03700000047683716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,5120,0.026196445027987163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,3072,0.011033777561452655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,4096,0.022060443957646687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,3584,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,5120,0.032702220810784235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,2560,0.009753777749008602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,4096,0.02800977892345852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,3072,0.009988444546858469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,2048,0.00817599975400501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,3072,0.01812622282240126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,3584,0.025245333711306255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,2560,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,2560,0.015847111741701763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,1536,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,3072,0.022993778189023335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,2048,0.008981333010726506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,2048,0.013746666411558786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,2560,0.020738666256268818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,1536,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,1536,0.011530666715568967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,2048,0.018571555614471436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,768,0.004591111093759537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,1536,0.016184000505341422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,1024,0.00796622203456031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,768,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,1024,0.014141332772043018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,512,0.003911111089918348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,768,0.013099555340078143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,512,0.007612444460391998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,256,0.003216888962520493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,512,0.007231111327807109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,512,0.012441777520709567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,256,0.006823110911581252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,1024,0.005423111220200856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,128,0.0029484445436133277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,256,0.011043555206722684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,1024,0.009662222531106737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,128,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,64,0.00282577777074443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,128,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,32,0.0028355556229750314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,128,0.009023110899660323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,768,0.008344000412358178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,65536,0.15510666370391846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,64,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,32,0.006737777756320105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,3584,0.02019466625319587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,65536,0.092013332578871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,65536,0.22871377733018664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,16384,0.0417431096235911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,51200,0.07409689161512586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,51200,0.17956177393595377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,12288,0.028100444210900202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,256,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,16384,0.025252444876564875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,65536,0.28643645180596244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,16384,0.06070311201943291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,10240,0.023826667004161414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,51200,0.22818756103515625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,12288,0.020612445142534044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,12288,0.04549955659442478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,16384,0.07987733019722833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,10240,0.018741332822375827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,10240,0.03928622272279527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,12288,0.06280977858437432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,8192,0.015200000670221118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,8192,0.03258488906754388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,10240,0.05409155620468987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,6144,0.015284443895022074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,51200,0.12132978439331055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,7168,0.013985777894655863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,7168,0.029342220889197454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,8192,0.04565599891874525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,5120,0.013088000317414602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,6144,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,6144,0.02571733295917511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,7168,0.04069955481423272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,5120,0.022514666120211285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,5120,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,6144,0.03615466753641764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,3584,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,5120,0.030969778696695965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,4096,0.018922666708628338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,8192,0.02016799979739719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,3584,0.010271111296282874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,3584,0.01717688971095615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,4096,0.02659911082850562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,7168,0.017472000585661996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,2560,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,3072,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,3584,0.024422221713595923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,3072,0.01574488812022739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,2048,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,2560,0.00908088849650489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,2560,0.013804444836245643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,3072,0.022257778379652236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,1536,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,2048,0.0120319997270902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,2048,0.00868622213602066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,4096,0.011180444724029966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,2560,0.020190222395790946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,1024,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,1536,0.01035377797153261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,2048,0.01793333391348521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,768,0.0041759999261962045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,4096,0.01054666688044866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,1024,0.007678222325113084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,1536,0.015509333875444202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,512,0.003543111185232798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,3072,0.009268444445398118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,768,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,1024,0.013735111388895245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,256,0.0032666667054096856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,512,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,768,0.0127688886390792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,512,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,256,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,256,0.006646222124497096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,512,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,128,0.003144888828198115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,128,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,64,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,128,0.006120000034570694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,32,0.002855111120475663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,1536,0.008223999705579545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,128,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,65536,0.13172977500491673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,1024,0.008642666869693333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,32,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,768,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,51200,0.09544710980521308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,65536,0.09077155590057373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,65536,0.18796355194515654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,16384,0.032495998673968844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,51200,0.07328710953394572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,51200,0.14912799994150797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,65536,0.27754934628804523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,12288,0.022667555345429316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,16384,0.024567110670937434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,256,0.010661332971519895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,51200,0.21886221567789713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,10240,0.017822222577200997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,12288,0.03870844509866502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,12288,0.020676443974177044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,16384,0.07733422517776489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,8192,0.01498044364982181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,64,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,10240,0.018631110588709515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,10240,0.03334577878316244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,12288,0.060108443101247154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,7168,0.013238222234778933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,8192,0.014987554815080432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,8192,0.0277502222193612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,6144,0.011704888608720569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,10240,0.052242666482925415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,7168,0.013949332965744866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,7168,0.025145777397685583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,5120,0.010205333431561789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,8192,0.04321066538492838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,6144,0.012808889150619507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,6144,0.022312889496485393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,7168,0.03911644551489089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,4096,0.008515555825498369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,5120,0.011668444507651858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,5120,0.01927911076280806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,6144,0.03387377659479777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,16384,0.049588445160124034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,3584,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,4096,0.010699555277824402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,3072,0.0070773329999711775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,4096,0.016210666961140104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,3584,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,5120,0.029891557163662378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,2560,0.006436444405052397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,3584,0.014859555496109856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,4096,0.02593244446648492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,3072,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,2048,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,3072,0.013448889056841532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,3584,0.023764444722069636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,2560,0.009248889154858058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,1536,0.004592000196377437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,3072,0.02142133315404256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,2048,0.00886222223440806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,2560,0.012067555553383298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,1024,0.0038888889054457345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,2048,0.010671111444632212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,2560,0.01960355540116628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,1536,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,768,0.0034959999223550162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,1536,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,2048,0.01756000022093455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,1024,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,512,0.0032044444233179092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,1024,0.00795377790927887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,768,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,1536,0.015154666370815702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,256,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,768,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,1024,0.013442666994200813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,512,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,128,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,512,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,256,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,768,0.012737777498033313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,64,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,256,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,128,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,32,0.0024968888610601425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,512,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,128,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,65536,0.0936631096733941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,256,0.01069866700304879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,64,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,51200,0.07318311267428927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,128,0.008681777450773451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,32,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,16384,0.02583644456333584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,65536,0.09133866760465835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,65536,0.18284532758924696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,51200,0.07353866762585111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,12288,0.02732533382044898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,51200,0.14358311229281956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,16384,0.023961777488390606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,65536,0.27326401074727374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,10240,0.023519999451107446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,12288,0.019718221492237516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,51200,0.21519645055135092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,12288,0.0380257765452067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,16384,0.07620177666346233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,8192,0.019183999962276883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,10240,0.017959111266665988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,10240,0.03261422117551168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,7168,0.01705422169632382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,8192,0.014505777094099255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,8192,0.026829333768950567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,10240,0.05096888873312208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,6144,0.014410666293568082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,8192,0.04257244533962674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,7168,0.02420444455411699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,6144,0.012449777788586087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,7168,0.0382008883688185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,6144,0.02177422245343526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,5120,0.012597333225939008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,5120,0.011408888631396823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,16384,0.0482666658030616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,6144,0.0331911113527086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,4096,0.008255110846625434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,5120,0.01868799991077847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,3584,0.0074746666683091064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,4096,0.010432889064153036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,5120,0.029263112280103896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,4096,0.01590044465329912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,3072,0.006790222393141852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,12288,0.05945689148373074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,3584,0.014613333675596448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,4096,0.025431111454963684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,2560,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,3072,0.00943288869327969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,3072,0.013259555730554791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,3584,0.023316444622145757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,2048,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,7168,0.01346577786737018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,2560,0.009036444127559662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,3072,0.02104088829623328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,1536,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,2048,0.008629333641793992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,2048,0.010399999717871347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,2560,0.019114666514926486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,1024,0.003617777592606015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,1536,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,1536,0.009268444445398118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,2048,0.01719555589887831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,768,0.0032711111836963226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,1024,0.007621333003044128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,1536,0.014828445182906257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,512,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,768,0.007256000406212277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,768,0.0069884442620807225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,3584,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,256,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,768,0.012068444656001197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,512,0.00665511108107037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,512,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,128,0.002266666748457485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,256,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,512,0.01107377807299296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,64,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,2560,0.011762667033407422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,128,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,32,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,128,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,256,0.01035733355416192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,65536,0.04985510971811083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,64,0.005542222410440445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,128,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,51200,0.041407111618253924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,65536,0.09075288640128241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,65536,0.14219466845194498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,1024,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,16384,0.016214221715927124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,51200,0.07291289170583089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,1024,0.01313688854376475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,51200,0.1107742190361023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,12288,0.01313244468635983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,65536,0.2702088885837131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,16384,0.03792266713248359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,10240,0.012091555529170565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,12288,0.01939733326435089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,51200,0.21250577767690024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,16384,0.07552533017264472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,12288,0.02986933456526862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,8192,0.012472000386979846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,10240,0.017906667457686532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,10240,0.025627555118666753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,12288,0.05849422348870171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,7168,0.011492444409264458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,8192,0.014490667316648694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,8192,0.021448888712459143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,10240,0.04998488889800178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,6144,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,32,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,8192,0.041562666495641075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,7168,0.013444444371594323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,5120,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,6144,0.012442666623327466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,6144,0.01779733267095354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,7168,0.037096887826919556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,4096,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,5120,0.011335999601417117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,5120,0.015537778536478678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,6144,0.03266577919324239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,16384,0.02313599983851115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,3584,0.006699555450015598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,4096,0.010275555153687796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,4096,0.01348000019788742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,5120,0.02898400028546651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,3072,0.006384000182151794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,3584,0.009711110757456886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,4096,0.025067556235525344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,2560,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,3072,0.011431111229790581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,3584,0.02289955483542548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,3072,0.0207368897067176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,2560,0.008973333570692275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,7168,0.01976355579164293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,2048,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,2560,0.010236444572607676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,1536,0.0041955556306574075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,2048,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,2560,0.0192577772670322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,2048,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,1024,0.0035084444615576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,1536,0.008267555799749162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,1536,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,2048,0.016883555385801528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,768,0.003236444460021125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,1024,0.0075137780772315125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,1024,0.007048889166778988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,1536,0.014529777897728814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,512,0.0028728888266616394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,3584,0.012434666355450949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,768,0.007270221908887227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,1024,0.01274400038851632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,256,0.002467555511328909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,3072,0.009305777649084726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,512,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,768,0.01179377817445331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,128,0.002208000048995018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,256,0.006695999867386288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,256,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,64,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,512,0.010953777366214328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,32,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,128,0.006572444405820634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,128,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,65536,0.04011822077963088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,256,0.01036622209681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,128,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,51200,0.032079110542933144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,65536,0.09070133500629002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,65536,0.14156444867451987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,51200,0.11051288578245376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,51200,0.07206222083833483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,16384,0.012109333442317115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,65536,0.27007465892367893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,512,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,12288,0.010119111173682744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,16384,0.023039999935362074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,51200,0.2138684458202786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,16384,0.07412533627616034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,12288,0.01925244430700938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,10240,0.00903555585278405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,64,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,12288,0.05823466512891981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,8192,0.008043555749787224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,32,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,10240,0.02536711096763611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,10240,0.017802667286660936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,7168,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,10240,0.04839555422465006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,8192,0.014196443888876172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,8192,0.021252445048756067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,6144,0.007119999991522894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,768,0.006549333532651265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,7168,0.013188444077968597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,8192,0.04109955496258206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,7168,0.019331556227472093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,5120,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,6144,0.012262221839692859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,6144,0.017284444636768766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,7168,0.036779556009504534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,16384,0.03756711218092177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,5120,0.011179555621412067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,5120,0.01553066737122006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,6144,0.03244888782501221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,12288,0.029576000240114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,3584,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,4096,0.01312711089849472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,3584,0.009854222337404886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,4096,0.024871110916137695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,3072,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,3584,0.012347555822796293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,3584,0.02269777821169959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,3072,0.009375111096435124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,2560,0.004982222285535601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,3072,0.011406222151385413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,2560,0.00903733323017756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,3072,0.020245333512624104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,2560,0.01034577770365609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,2048,0.004592000196377437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,4096,0.006888000087605582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,1536,0.00387999994887246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,2048,0.009108444054921469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,5120,0.028806222809685603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,1536,0.008073777788215214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,1024,0.003216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,1536,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,1536,0.01442399952146742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,1024,0.0070062221752272705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,1024,0.007345777418878343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,768,0.002969777832428614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,1024,0.0127688886390792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,768,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,512,0.0026595555245876312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,768,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,2560,0.01891911029815674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,256,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,512,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,4096,0.010239111052619087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,256,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,128,0.0022008888837363985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,256,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,256,0.010070222119490305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,128,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,64,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,128,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,32,0.0021128888345426982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,128,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,65536,0.03759110967318217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,64,0.00499911109606425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,51200,0.032138665517171226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,32,0.0052977779673205475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,768,0.011409777734014722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,512,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,65536,0.14085066318511963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,12288,0.010011555420027839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,10240,0.009091555244392818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,51200,0.10996533102459377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,2048,0.01679999960793389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,16384,0.03717511230044895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,512,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,7168,0.007897777689827813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,6144,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,12288,0.029176887538697984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,5120,0.006882666713661618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,8192,0.021310221817758348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,4096,0.0063484443558586975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,6144,0.01752088963985443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,3584,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,5120,0.015513777732849121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,3072,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,4096,0.013412444127930535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,2560,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,3584,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,2048,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,16384,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,3072,0.011365333365069495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,1536,0.0038275557259718576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,1024,0.0031511110977994073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,2560,0.0101742222905159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,768,0.0028124445428450904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,2048,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,512,0.0025128889828920364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,8192,0.008260444634490544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,1536,0.008226667013433244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,256,0.002193777718477779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,1024,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,768,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,64,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,10240,0.025058666865030926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,512,0.005783111270931032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,32,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,256,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,7168,0.019327110714382596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,65536,0.035813334915373064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,51200,0.02832888894610935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,16384,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,64,0.005024888863166173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,12288,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,32,0.004953777624501122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,10240,0.014813333749771118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,51200,0.10991822348700629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,8192,0.012903999951150684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,16384,0.036995556619432234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,12288,0.029159112109078303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,2048,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,7168,0.012039110892348819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,10240,0.025109334124459162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,6144,0.01068800025516086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,5120,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,8192,0.021140444609853957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,4096,0.008959111240175035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,7168,0.01925066610177358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,6144,0.01756355497572157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,3584,0.008032889001899296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,128,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,5120,0.015458666616015963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,2560,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,4096,0.013350221845838757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,2048,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,3584,0.012306667036480375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,3072,0.011422221859296164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,128,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,1024,0.003144888828198115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,2560,0.010225777824719747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,768,0.002809777855873108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,2048,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,512,0.002481777841846148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,1536,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,65536,0.14067910777197942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,256,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,1024,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,128,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,64,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,768,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,512,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,256,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,128,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,64,0.005006222261322869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,32,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,3072,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,1536,0.0035982223020659555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,16384,0.5807271003723145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,12288,1.3722737630208333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,16384,1.777963638305664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,12288,0.44624890221489805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,12288,1.3474133809407551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,16384,1.8716729482014973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,10240,0.3833173380957709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,10240,1.0946132871839735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,8192,0.7861893441942004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,8192,0.8470097647772895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,51200,2.139968024359809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,8192,0.30502933926052517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,7168,0.6840711169772677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,7168,0.2782319916619195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,7168,0.8031022283766004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,6144,0.6559031274583604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,6144,0.23862756623162162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,6144,0.6740258004930285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,5120,0.518072870042589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,10240,1.124378628200955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,5120,0.23084266980489096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,4096,0.3914817704094781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,4096,0.17033333248562285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,4096,0.43866221110026044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,5120,0.5305217636956109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,3584,0.3433510992262099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,3584,0.153857774204678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,3072,0.320632881588406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,2560,0.24272354443868002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,3072,0.13219911522335476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,3584,0.37943198945787215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,51200,5.472095065646702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,3072,0.31918401188320583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,2048,0.2027075555589464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,2560,0.12194311618804932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,2560,0.28087109989590114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,1536,0.1650302277670966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,2048,0.09913244512346055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,1024,0.11793511443667942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,768,0.09246666563881768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,1536,0.17442222436269125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,1536,0.07751999961005317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,1024,0.13324356079101562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,512,0.07291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,768,0.055283556381861367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,256,0.04379466507169935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,512,0.051051557064056396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,768,0.1066266695658366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,512,0.09361777702967326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,256,0.04706577791108025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,256,0.08291200134489271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,64,0.03430755601988898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,128,0.04556711183653938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,32,0.03770577907562256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,51200,6.492204454210069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,128,0.07794577545589872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,64,0.07756355735990736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,32,0.07800889015197754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,2048,0.2206044461992052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,1024,0.061866667535569936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,128,0.03873599900139703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,16384,1.4694444868299696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,51200,1.5728719499376085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,65536,2.1773315005832248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,16384,0.5582266913519965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,16384,1.468303150600857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,51200,4.028612348768446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,12288,1.0894621743096244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,12288,0.35773245493570965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,12288,1.0375137329101562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,65536,5.306450737847222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,10240,0.930572403801812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,10240,0.3358248869578044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,51200,5.047413296169705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,8192,0.7517937554253472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,10240,0.8543510966830783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,7168,0.21408977773454455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,7168,0.6798506842719184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,8192,0.670563538869222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,8192,0.243867556254069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,7168,0.599364439646403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,6144,0.6397626664903429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,6144,0.4897644254896376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,6144,0.1856853299670749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,5120,0.16122044457329646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,5120,0.4614773326449924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,4096,0.3273084428575304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,5120,0.41427111625671387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,4096,0.13472178247239855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,3584,0.301512877146403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,3072,0.26700178782145184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,3584,0.1338826682832506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,4096,0.41188799010382754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,3584,0.29165156682332355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,2560,0.22993689113193086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,3072,0.10601955652236938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,65536,7.003838433159722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,2048,0.16680088308122423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,3072,0.24875911076863608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,2560,0.11816088358561198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,1536,0.13845600022210017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,1024,0.0811777777141995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,1536,0.061031997203826904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,2560,0.25764621628655326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,768,0.06799199846055773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,1024,0.049068444304996066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,1536,0.14428444703420004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,1024,0.1191671159532335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,768,0.048283557097117104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,256,0.039445334010654025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,768,0.0856319997045729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,512,0.07305955224566989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,128,0.030066665675905015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,256,0.03861422340075175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,256,0.06647466950946383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,64,0.030714667505688135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,32,0.031182222896152075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,128,0.037083556254704796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,128,0.06217422087987264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,64,0.062337776025136314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,32,0.06224711073769463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,2048,0.08826311429341634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,2048,0.17460978031158447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,512,0.05270666546291775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,512,0.04189155499140421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,16384,0.3849297894371881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,65536,0.6746222178141276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,51200,0.5325724283854166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,16384,0.16103733910454646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,16384,0.4565484258863661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,65536,1.7447990841335719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,51200,1.5132915708753798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,12288,0.398693323135376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,12288,0.12292799684736465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,12288,0.37524088223775226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,8192,0.2599173386891683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,10240,0.10427644517686631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,51200,1.5820195939805772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,65536,2.09426392449273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,8192,0.08426222536298965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,10240,0.28446756468878853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,7168,0.2268444432152642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,8192,0.21965511639912924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,6144,0.16059554947747126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,6144,0.0649777783287896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,5120,0.1507564385732015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,7168,0.19717866844601104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,5120,0.05739911397298177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,4096,0.1171591149436103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,5120,0.13975466622246638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,3584,0.09551644325256348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,4096,0.047127111090554126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,3072,0.08618399831983779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,4096,0.11582577228546143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,3584,0.04225511021084256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,3584,0.09998666577868992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,10240,0.2503653367360433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,3072,0.037610666619406805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,2048,0.0670657753944397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,3072,0.08686221970452203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,2560,0.03324355681737264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,1536,0.05222488774193657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,2560,0.07549599806467693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,2048,0.06214577621883816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,1536,0.02370577719476488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,1024,0.03256533212131924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,7168,0.07455288701587252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,1536,0.04976177877850003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,768,0.02808800008561876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,1024,0.01976977785428365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,512,0.0196079992585712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,6144,0.17033067014482287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,768,0.03125244379043579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,256,0.014324444863531323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,512,0.01646933290693495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,512,0.02569155560599433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,128,0.011703111231327057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,2560,0.07805244127909343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,256,0.022892443670166865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,64,0.010009778042634329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,128,0.015049777097172208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,32,0.01033688916100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,2048,0.02879377868440416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,128,0.022030222747060988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,64,0.02204977803760105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,32,0.02201333310869005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,1024,0.03808888792991638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,256,0.015112888481881885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,65536,0.48558134502834743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,65536,1.2009333504570854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,51200,0.9499724706013998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,16384,0.36086132791307235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,51200,0.38974934154086643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,12288,0.2591448889838325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,65536,1.5029288397894964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,16384,0.12672889232635498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,768,0.018023111753993563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,16384,0.35929687817891437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,10240,0.21276888582441542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,12288,0.09312799904081558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,51200,1.1690026389227974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,10240,0.0819751090473599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,12288,0.33866222699483234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,8192,0.15647644466824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,7168,0.15191733837127686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,8192,0.06919111145867242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,7168,0.05791644255320231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,8192,0.16912799411349824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,5120,0.10858133104112412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,6144,0.051226665576299034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,7168,0.14785155985090467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,4096,0.09054844246970283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,3584,0.07780177725685967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,5120,0.1087484425968594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,6144,0.12866577837202284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,5120,0.04332000017166138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,4096,0.09335644377602471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,4096,0.03627999954753452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,3072,0.06538666619194879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,2560,0.054455998871061534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,3584,0.03229422370592753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,2048,0.045160889625549316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,3584,0.07805599768956502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,2560,0.026338666677474976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,3072,0.06782311201095581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,2560,0.05901422103246053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,2048,0.022663111488024395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,1024,0.02419111132621765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,2048,0.04861866765552097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,1536,0.019711999429596793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,768,0.01885600056913164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,6144,0.1283751063876682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,1024,0.016378665963808697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,512,0.013750221994188098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,768,0.015152888165579902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,768,0.02522755497031742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,256,0.010594666832023196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,512,0.013844444519943662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,512,0.020741333564122517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,128,0.008974221845467886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,256,0.012784889174832238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,256,0.018360000517633226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,64,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,128,0.012812444733248817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,3072,0.02983022067281935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,32,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,128,0.01788444485929277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,1536,0.03456266721089681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,64,0.017916444275114272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,32,0.01757066614098019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,10240,0.2078782187567817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,1024,0.030176000462638006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,65536,0.44426311386956113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,65536,1.0029964447021484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,51200,0.7472151120503744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,51200,0.34437955750359434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,12288,0.19707111517588297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,16384,0.10657422410117255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,65536,1.3115261925591362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,16384,0.31620799170600045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,51200,1.068932427300347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,10240,0.20116799407535127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,12288,0.2173546685112847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,12288,0.08590400218963623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,8192,0.16178399986690944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,1536,0.03919733232922024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,10240,0.07349689139260186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,8192,0.05769777960247464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,7168,0.14302755726708308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,10240,0.1869164440366957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,7168,0.0511448880036672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,6144,0.12088621987236871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,5120,0.10525866349538167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,7168,0.12607644663916692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,6144,0.04391377833154467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,16384,0.2899102105034722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,4096,0.06832533412509494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,6144,0.14095732900831434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,3584,0.058802664279937744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,5120,0.09392888678444757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,4096,0.031238221459918555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,3072,0.06064889166090223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,4096,0.07521510786480375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,2560,0.04810933272043864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,3072,0.026530666483773127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,3584,0.06640266709857516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,3072,0.057413332992129855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,2048,0.03921422362327576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,2560,0.023678221636348303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,1536,0.03129955463939243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,2560,0.04982666505707634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,2048,0.020791111720932853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,1024,0.020377778344684176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,2048,0.04172711239920723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,1536,0.017502221796247695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,768,0.016304888659053378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,8192,0.14413689242468938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,1536,0.03416444526778327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,1024,0.014879999889267815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,512,0.01275199982855055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,1024,0.02630133264594608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,768,0.013629333012633853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,512,0.012706666356987424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,768,0.02200444373819563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,5120,0.03760622276200189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,512,0.017904000149832833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,128,0.00813600007030699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,64,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,256,0.011763555308183035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,32,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,3584,0.028598222467634413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,128,0.011894221935007306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,128,0.015652444627549913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,64,0.0155102229780621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,32,0.015577778220176697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,65536,0.35853155453999835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,51200,0.2798124419318305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,51200,0.6339431338840061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,65536,0.8847466574774848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,256,0.009000889129108852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,16384,0.24186311827765572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,12288,0.18573421902126738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,51200,0.7894150945875379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,16384,0.09475111299090916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,65536,1.0162898169623482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,256,0.016164445214801364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,16384,0.23847733603583443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,10240,0.16898932721879747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,12288,0.17268889480166963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,8192,0.13267378012339273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,10240,0.05858133236567179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,7168,0.11755911509195964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,8192,0.047621332936816745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,6144,0.09391999906963772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,10240,0.15029866165584987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,7168,0.04867022236188253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,8192,0.11604089207119411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,5120,0.07350844144821167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,7168,0.10390044583214654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,6144,0.03715022073851691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,4096,0.06863466898600261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,5120,0.03453866640726725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,3584,0.06110577450858223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,6144,0.09655555751588608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,4096,0.02737955583466424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,3072,0.053540445036358304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,5120,0.07773600021998088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,3584,0.027254222167862788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,4096,0.0624177787039015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,3072,0.023223999473783705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,3584,0.05495199892255995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,2048,0.03546311126814948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,2560,0.021578666236665513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,1536,0.027645332945717707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,3072,0.04913155568970574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,2560,0.04243377844492594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,2048,0.01834488908449809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,1536,0.01614755557643043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,2048,0.03579377796914842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,768,0.014763555592960782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,12288,0.0701191094186571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,1536,0.02899466620551215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,1024,0.01346933344999949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,512,0.010855110983053843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,1024,0.02224622170130412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,256,0.008174222376611497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,768,0.01240533341964086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,768,0.01863466699918111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,512,0.011622221933470832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,128,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,256,0.01090666651725769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,512,0.015442666080262927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,256,0.013908444179428948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,64,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,32,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,128,0.01349422252840466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,64,0.013455111119482251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,2560,0.044277333550983004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,32,0.013459555804729462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,1024,0.016882666283183627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,65536,0.3136168850792779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,65536,0.8100133472018771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,51200,0.6032097604539659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,51200,0.24740621778700086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,16384,0.22887555758158365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,65536,0.8864951133728027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,51200,0.6952382193671333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,16384,0.07767822345097859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,12288,0.1689795520570543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,128,0.010718222293588849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,16384,0.2312328815460205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,10240,0.1390142175886366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,12288,0.06251911322275798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,8192,0.09963288572099473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,10240,0.05751111110051473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,7168,0.09629778067270915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,12288,0.1734622187084622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,8192,0.04287999868392944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,10240,0.1335502200656467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,7168,0.037674665451049805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,8192,0.10424000024795532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,5120,0.0665502217080858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,6144,0.03443288803100586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,5120,0.029303110308117334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,6144,0.0851679974132114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,3584,0.04831822382079231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,5120,0.06906221972571479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,4096,0.025518221987618342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,3072,0.04179733329349094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,4096,0.0568373335732354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,3584,0.023028444912698533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,2560,0.03279555506176419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,3584,0.05019022358788384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,2048,0.02846577763557434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,3072,0.021186666356192693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,3072,0.043819556633631386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,2560,0.018985778093338013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,1536,0.02278577784697215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,2560,0.038098666403028704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,2048,0.016700444949997794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,1024,0.015831111205948722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,2048,0.031975110371907554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,1536,0.01465777721669939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,1536,0.026157332791222468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,1024,0.012420444852775998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,512,0.010152888794740042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,1024,0.01999022232161628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,6144,0.07804622252782185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,768,0.011420444481902652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,256,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,512,0.010728889041476779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,768,0.01703999936580658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,4096,0.05482755435837639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,128,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,512,0.014144000079896716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,256,0.010232888989978367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,64,0.00553955551650789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,32,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,128,0.010213333699438307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,256,0.013018666870064206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,128,0.01237777786122428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,64,0.012376888758606382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,32,0.012444444828563266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,65536,0.26525600751241046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,65536,0.5868142445882162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,51200,0.2063902219136556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,51200,0.5081297556559244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,768,0.013146667016877068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,16384,0.16580533981323242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,12288,0.11249244213104248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,16384,0.06538044744067721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,65536,0.7809928788079156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,10240,0.11525332927703857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,51200,0.5977190865410699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,7168,0.09439644548628066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,16384,0.20926755004458955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,12288,0.1344755490620931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,8192,0.08146755562888251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,10240,0.04428889022933113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,7168,0.06940088669459026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,6144,0.06411733229955037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,10240,0.11475200123257107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,7168,0.03702311052216424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,8192,0.09202755822075738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,5120,0.054153776831097074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,7168,0.08112711376614042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,6144,0.02871911062134637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,4096,0.0418524444103241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,5120,0.02721066607369317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,3584,0.03831555445988973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,5120,0.060079110993279346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,4096,0.02159999973244137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,4096,0.04970488945643107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,3072,0.0322746667597029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,3584,0.01996799972322252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,2560,0.026520000563727483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,3072,0.019034667147530448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,3584,0.04376888937420315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,2048,0.023327999644809302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,3072,0.03906666570239597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,12288,0.05166133244832357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,2560,0.01697422232892778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,1536,0.018096889058748882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,2560,0.03375733229849074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,2048,0.015047111445003085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,1024,0.013838222457302941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,1536,0.013136000268989138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,768,0.011344888971911537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,8192,0.03501244386037191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,1536,0.023502222365803186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,1024,0.011357333097192975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,512,0.008631111019187504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,1024,0.018033777674039204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,768,0.010616888602574667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,256,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,512,0.010012444522645738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,768,0.01534844438234965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,128,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,6144,0.07134044170379639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,512,0.013002666334311167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,64,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,256,0.009387555221716562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,32,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,128,0.009328888522254096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,256,0.011927111281288994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,128,0.011360888679822287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,64,0.011355555719799466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,32,0.011417778001891242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,65536,0.45151646931966144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,65536,0.24260799090067545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,51200,0.4139848815070258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,16384,0.1211235523223877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,51200,0.19125066863165963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,2048,0.028761777612898085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,12288,0.09903289212120904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,65536,0.6584373580084907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,10240,0.08113955789142184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,16384,0.15192355049981013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,51200,0.5083537631564671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,8192,0.07365777757432726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,10240,0.04582755433188545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,7168,0.058168000645107694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,10240,0.10929955376519097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,8192,0.03347288899951511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,8192,0.08015555805630155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,6144,0.05696800020005968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,5120,0.047169778082105845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,7168,0.030926220946841772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,6144,0.027679110566775005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,7168,0.07242755757437812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,4096,0.03720888826582167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,5120,0.024433778391944036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,6144,0.06087911128997803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,3584,0.03234488765398661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,5120,0.05276088913281759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,4096,0.021307556165589228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,3072,0.028355555401908025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,3584,0.01956266661485036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,4096,0.04347199863857693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,2560,0.02346755564212799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,3072,0.017872888180944655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,16384,0.06170044342676798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,2560,0.016201777590645682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,3072,0.03461777832772996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,1536,0.015091554986106025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,2560,0.030398223135206435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,12288,0.11632977591620551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,2048,0.014490667316648694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,1024,0.011643555429246692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,2048,0.02571022179391649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,1536,0.01275466630856196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,768,0.008971555365456475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,1536,0.0206977774699529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,512,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,1024,0.015998222761683993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,768,0.01033688916100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,256,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,768,0.013795555465751223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,512,0.009655999640623728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,128,0.0053119998839166425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,512,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,256,0.009288888838556077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,256,0.010680889089902243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,64,0.004534222185611725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,128,0.009364444348547194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,32,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,3584,0.038466665479871966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,128,0.010316444767846001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,2048,0.019780443774329293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,64,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,12288,0.04885244369506836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,32,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,65536,0.4552035331726074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,51200,0.34373066160413956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,1024,0.0111191107167138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,51200,0.15374844604068333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,65536,0.19749600357479521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,16384,0.11205600367652045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,12288,0.08411288923687404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,16384,0.04966400067011515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,10240,0.08072000079684787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,51200,0.420214229159885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,12288,0.0520977775255839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,16384,0.128492448065016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,8192,0.06502044200897217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,10240,0.037485334608289934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,7168,0.057987557517157666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,12288,0.09707999891704983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,8192,0.030753778086768255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,10240,0.08297777838177152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,6144,0.04684888985421923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,8192,0.0909608867433336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,5120,0.040780444939931236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,6144,0.023416888382699754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,7168,0.06198044617970785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,4096,0.03159822358025445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,5120,0.02079288827048408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,6144,0.05210399958822462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,3584,0.026477333572175767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,5120,0.05927111042870415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,3072,0.02260977857642704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,4096,0.018062222335073683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,4096,0.03772266705830892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,2560,0.020397333635224234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,3584,0.0336426662074195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,2048,0.016533333394262526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,3072,0.015519110692871941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,2560,0.0138568886452251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,1536,0.01351288871632682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,65536,0.5391875372992622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,2560,0.02644799980852339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,1024,0.010098666780524785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,2048,0.022517333428064983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,1536,0.018256889449225534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,768,0.008403555386596257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,1024,0.00998755544424057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,1024,0.013960889644092984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,512,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,768,0.009313777916961247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,768,0.012456888953844706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,256,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,512,0.008683555656009251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,3584,0.01647911138004727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,256,0.008519111408127679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,128,0.0042160000238153665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,256,0.009679111341635386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,3072,0.03069066670205858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,64,0.0038968887594011095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,32,0.003891555385457145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,128,0.008375999828179678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,2048,0.01279022213485506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,128,0.009310222334331935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,64,0.009343110852771336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,32,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,1536,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,7168,0.026271111435360376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,65536,0.18839289082421196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,51200,0.33372799555460614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,16384,0.11093422439363267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,12288,0.085043556160397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,65536,0.47848886913723415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,51200,0.1477173301908705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,16384,0.04837244417932299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,512,0.010543111297819348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,10240,0.06635110908084445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,51200,0.3806355529361301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,12288,0.038712001509136625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,8192,0.04831822382079231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,16384,0.11400889025794135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,7168,0.04651822315322029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,10240,0.0900497767660353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,6144,0.042599111795425415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,12288,0.08682399988174438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,7168,0.02587555514441596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,8192,0.060384001996782094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,5120,0.03424177898301019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,7168,0.053257776631249316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,6144,0.02273422148492601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,6144,0.04758310980266995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,4096,0.026545777916908264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,65536,0.41980889108445907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,5120,0.0203813330994712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,3584,0.025044444534513686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,4096,0.017323555217848886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,3072,0.021107556091414556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,3584,0.016114667057991028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,5120,0.04021688964631822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,2560,0.018409777018758986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,4096,0.03326933251486884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,3072,0.014765333798196582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,3072,0.02630311085118188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,2048,0.015385778413878547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,3584,0.030167112747828167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,1536,0.012786666552225748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,2560,0.013761777844693927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,8192,0.027815110153622095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,2048,0.012447111308574677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,1024,0.009456000394291347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,2048,0.01960711181163788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,768,0.007732444339328342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,2560,0.02348088887002733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,1024,0.009720889230569204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,512,0.006485333459244833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,1536,0.01110755569405026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,1024,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,256,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,768,0.009332444104883406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,768,0.011396444506115384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,1536,0.0162462227874332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,512,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,512,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,256,0.008351999852392409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,64,0.003918222255176968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,32,0.0042035554846127825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,128,0.008280889027648503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,64,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,32,0.008640000389681922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,128,0.008760000268618265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,10240,0.03412177827623155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,51200,0.2164391146765815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,65536,0.15359466605716282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,16384,0.08789155880610149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,65536,0.41779200236002606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,51200,0.33337065908643937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,12288,0.06649600134955512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,51200,0.12348088953230117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,128,0.0042160000238153665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,10240,0.05300177799330818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,16384,0.10105599959691365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,256,0.00908622228437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,12288,0.03488444288571676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,8192,0.045668443044026695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,12288,0.07765422264734904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,7168,0.03914666506979201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,10240,0.06552088922924466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,8192,0.022781333989567224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,65536,0.34337155024210614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,6144,0.03402488761478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,7168,0.020766221814685397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,5120,0.028120001157124836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,7168,0.04807288779152764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,6144,0.01862844493654039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,4096,0.02348711093266805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,6144,0.04212444358401828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,5120,0.01687733332316081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,3584,0.020430222153663635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,5120,0.036031110419167414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,3072,0.016862221890025668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,4096,0.015205333630243937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,4096,0.030360887447992962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,16384,0.039834668238957725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,2560,0.015028445257080926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,3584,0.013530666629473368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,2048,0.012633778154850006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,3072,0.012842666771676807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,2560,0.01143022212717268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,3072,0.02421422302722931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,1536,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,2048,0.010317333042621613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,2560,0.02097333305411869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,10240,0.03096888793839349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,2048,0.017847999930381775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,1536,0.009724444813198512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,768,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,1024,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,512,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,1024,0.011744000017642975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,768,0.008378666308191087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,256,0.004393777913517422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,768,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,8192,0.05390488770272997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,512,0.008040889269775814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,128,0.0038906666967603895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,512,0.009045333498054082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,256,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,256,0.008672888908121321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,128,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,32,0.0035173334181308746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,128,0.008340444829728868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,64,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,32,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,3584,0.027336888843112524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,65536,0.2966008981068929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,51200,0.1965386602613661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,1024,0.007928888830873702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,65536,0.1447546614540948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,65536,0.3649262322319879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,16384,0.07772533098856609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,1536,0.014858666393491956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,12288,0.058711998992496066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,51200,0.1133306688732571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,16384,0.038005332152048744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,10240,0.0459191103776296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,12288,0.0303146673573388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,16384,0.09011110994550918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,8192,0.03794844282997979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,12288,0.06846221950319079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,10240,0.02717422280046675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,64,0.00350222239891688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,10240,0.05847999784681532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,8192,0.021897777915000916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,6144,0.02884711159600152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,7168,0.020412445068359375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,8192,0.04810133245256212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,5120,0.023815110325813293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,7168,0.04302666584650675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,4096,0.019542222221692402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,5120,0.016304888659053378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,5120,0.03240177697605557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,3584,0.0170817772547404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,4096,0.014312888185183207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,4096,0.027245332797368366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,3072,0.015360888507631091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,3584,0.013229333692126803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,3584,0.024541333317756653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,51200,0.3181031015184191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,2560,0.01314755529165268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,3072,0.012372444073359171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,2048,0.011323555476135679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,3072,0.021910222040282354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,2560,0.011342222491900126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,1536,0.009119110802809397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,7168,0.031397332747777305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,2048,0.010159999959998662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,1024,0.0070471109615431885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,2048,0.016366221838527255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,1536,0.00943199959066179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,768,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,1536,0.013774221969975365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,1024,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,6144,0.01833777791923947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,1024,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,6144,0.03782577647103204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,768,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,256,0.004219555606444676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,768,0.009720889230569204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,512,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,128,0.0035093331502543557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,512,0.008464888566070134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,64,0.0032044444233179092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,128,0.007487110793590546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,256,0.008096888661384583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,32,0.003506666670242945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,128,0.00812266684240765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,64,0.007664889097213745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,65536,0.24345689349704316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,32,0.007736889024575551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,2560,0.01885955532391866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,51200,0.18414666917588976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,65536,0.13777600394354925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,65536,0.3133617771996392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,16384,0.058211558394961886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,12288,0.04387199878692627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,51200,0.1077386670642429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,16384,0.0365955564710829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,512,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,10240,0.035895999934938215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,12288,0.03554844525125291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,8192,0.029669334491093952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,10240,0.027063111464182537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,12288,0.05908799833721585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,7168,0.025799110531806946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,8192,0.021135111649831135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,256,0.00756888919406467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,10240,0.05148799882994758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,6144,0.022281777527597215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,8192,0.04458489020665487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,5120,0.018183110488785636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,7168,0.019282667173279654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,6144,0.01756088932355245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,4096,0.015504888362354703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,5120,0.015776889191733468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,3584,0.013949332965744866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,6144,0.033555554019080266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,5120,0.029397332006030615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,3072,0.012142221960756512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,4096,0.024056888288921777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,3584,0.01275644451379776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,2560,0.010871999793582492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,3584,0.02181688944498698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,3072,0.012083555261294046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,3072,0.019500444332758587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,2048,0.008999110923873054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,51200,0.24578844176398384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,2560,0.017166222135225933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,16384,0.07869333028793335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,1536,0.007465778125656976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,2048,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,2048,0.014668444792429606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,1024,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,1536,0.009365333451165093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,768,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,1536,0.012418666647540199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,1024,0.008628444539176093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,512,0.004229333251714706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,1024,0.010013333625263637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,7168,0.037653333610958524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,768,0.008211555580298105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,256,0.0035075553589397003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,768,0.008957333034939235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,512,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,4096,0.013814222481515674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,512,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,64,0.002814222127199173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,256,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,128,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,32,0.0031591111587153543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,128,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,64,0.0072266666425599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,65536,0.21736444367302787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,32,0.007264888948864407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,51200,0.15543911192152235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,65536,0.09473777479595608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,65536,0.2644142309824626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,16384,0.05617066886689928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,51200,0.07759466436174181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,51200,0.20498932732476127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,12288,0.04137244489457872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,16384,0.025176000263955858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,10240,0.03339555528428819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,16384,0.06702044275071886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,12288,0.021311110920376245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,128,0.003133333391613431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,8192,0.026456000076399908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,256,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,10240,0.021558221843507554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,7168,0.02420888841152191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,10240,0.044194665220048696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,8192,0.015666666958067153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,6144,0.020378667447302077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,8192,0.0362737774848938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,7168,0.014290667242474027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,7168,0.0329093337059021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,6144,0.013182222015327878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,2560,0.01088177743885252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,4096,0.01484266585773892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,6144,0.029084444046020508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,3584,0.013442666994200813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,5120,0.012151999606026543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,5120,0.024791111548741657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,3072,0.012063999970753988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,4096,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,3584,0.010308444499969482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,2560,0.010630221830474006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,4096,0.020485333270496793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,3072,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,3584,0.019270223047998216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,2048,0.008392000363932716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,2560,0.009369778136412302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,3072,0.01711466742886437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,2048,0.009008888569143083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,2560,0.01482488877243466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,1024,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,2048,0.013065777719020844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,768,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,1536,0.01108888867828581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,1024,0.007844444778230455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,512,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,1024,0.009204444785912832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,768,0.007700444095664554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,256,0.003653333418899112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,768,0.008183110919263627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,512,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,128,0.003169777699642711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,5120,0.01759466694460975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,512,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,256,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,256,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,128,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,32,0.0029564443975687027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,128,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,64,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,65536,0.1280719969007704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,32,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,1536,0.007066667079925537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,65536,0.08643555641174316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,16384,0.03485511077774896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,1536,0.008310221963458592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,51200,0.06957688596513537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,12288,0.024733333124054804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,12288,0.051367110676235624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,51200,0.16589866744147405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,16384,0.023959111836221483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,16384,0.055236445532904736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,12288,0.020062221421135795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,12288,0.04260266820589701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,8192,0.016539555456903245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,64,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,10240,0.018216888109842937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,10240,0.03632355398601956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,7168,0.014927999840842353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,8192,0.014845333165592618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,6144,0.012800888882742988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,8192,0.03054311209254795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,7168,0.013453333742088742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,5120,0.011065777805116443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,7168,0.02735288937886556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,51200,0.09765511088901097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,6144,0.012496000362767113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,6144,0.024102222588327196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,5120,0.011417778001891242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,3584,0.008344000412358178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,5120,0.020625778370433383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,4096,0.010417777631017896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,3072,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,4096,0.017628444565667045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,3584,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,3584,0.01614755557643043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,10240,0.020273778173658583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,3072,0.009480888644854227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,2048,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,3072,0.014654222461912366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,2560,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,1536,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,2560,0.01278222186697854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,1024,0.004223111189073986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,2048,0.00865955568022198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,2048,0.011714666253990598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,768,0.0035751110149754416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,1536,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,1024,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,1536,0.0099982221921285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,512,0.003208000005947219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,768,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,1024,0.008037333687146505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,256,0.002834666727317704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,768,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,512,0.0069857777820693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,128,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,256,0.006664888726340399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,512,0.00665155549844106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,64,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,128,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,65536,0.21243911319308809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,128,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,65536,0.09864355458153619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,2560,0.0068693334857622785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,64,0.006082666830884085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,32,0.006228444476922353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,51200,0.0850728882683648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,65536,0.0862391127480401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,16384,0.03269422385427687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,51200,0.06952266560660468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,65536,0.17275644673241508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,12288,0.024599111742443506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,51200,0.136245330174764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,16384,0.023414222730530634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,10240,0.02107644412252638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,16384,0.04582488867971632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,12288,0.0196782218085395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,8192,0.017198221551047433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,12288,0.035931554105546736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,10240,0.01796888808409373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,7168,0.015881778465376962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,8192,0.014644443988800049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,10240,0.03084711233774821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,4096,0.009483555952707926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,6144,0.013829333086808523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,8192,0.025832888152864244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,7168,0.013564444250530668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,5120,0.011983111500740051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,4096,0.010315555665228102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,32,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,256,0.006292444550328785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,3584,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,5120,0.011438222395049201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,6144,0.020440000626775954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,4096,0.010330666270520952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,5120,0.017860444055663217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,4096,0.015184000134468079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,3584,0.009973333113723332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,2560,0.00742222203148736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,3072,0.009490666290124258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,2048,0.0063031112982167145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,3072,0.012602667013804117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,2560,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,2560,0.011122666299343109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,1536,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,2048,0.008621333373917473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,2048,0.010280888941552904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,1536,0.00814400033818351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,1536,0.008773333496517604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,1024,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,1024,0.007553777760929531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,768,0.0042275554604000514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,7168,0.023186667097939387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,6144,0.0124177775449223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,768,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,768,0.006694222076071634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,512,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,512,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,3584,0.013833777772055732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,128,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,256,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,256,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,64,0.002458666761716207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,128,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,1024,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,32,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,65536,0.08642399973339504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,64,0.005541333307822545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,32,0.005856888989607493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,65536,0.08588355779647827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,51200,0.06957955492867364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,512,0.0034933334423436057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,65536,0.1654871172375149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,51200,0.06926133235295613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,16384,0.024884444144037034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,3072,0.008276444342401292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,256,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,51200,0.1305039988623725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,16384,0.022801778382725183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,10240,0.018446221947669983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,16384,0.044063998593224414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,12288,0.01906133360332913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,8192,0.015368888775507608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,12288,0.034937779108683266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,10240,0.017543110582563613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,7168,0.015953777564896476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,8192,0.014210666219393412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,10240,0.029562665356530085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,128,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,6144,0.01366133325629764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,8192,0.02463200026088291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,7168,0.013224889006879596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,4096,0.009855111440022787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,7168,0.022337777747048274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,6144,0.012106666962305704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,5120,0.011324444578753578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,3584,0.007052444749408298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,6144,0.019727110862731934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,3072,0.006814222368929122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,5120,0.017456000049908955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,4096,0.009993777506881291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,2560,0.005879111174080107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,4096,0.014961777461899651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,3072,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,2048,0.004971555537647671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,3072,0.012463111016485425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,2560,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,12288,0.020782222350438435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,1536,0.00423555572827657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,2048,0.008345777789751688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,1024,0.003528000166018804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,2048,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,1536,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,768,0.0032035555276605818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,1536,0.008642666869693333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,1024,0.00755466686354743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,512,0.0028862222615215513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,1024,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,768,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,5120,0.013231111069520315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,256,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,512,0.006908444480763541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,128,0.0024213333510690266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,512,0.006107555495368109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,256,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,64,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,256,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,128,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,3584,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,32,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,128,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,3584,0.013757333159446716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,65536,0.04577955603599548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,2560,0.008672888908121321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,64,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,32,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,65536,0.08531200223498875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,65536,0.12501955032348633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,51200,0.06835644774966769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,12288,0.011530666715568967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,16384,0.02216533323129018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,51200,0.0982675552368164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,10240,0.012476444244384766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,16384,0.03424355718824599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,12288,0.01888444523016612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,8192,0.012063110868136087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,10240,0.01741066740618812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,12288,0.027132444911532935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,10240,0.023353776997990076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,7168,0.011038222246699862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,8192,0.014149333039919535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,768,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,6144,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,8192,0.019669334093729656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,5120,0.008375110725561777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,7168,0.013079110946920184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,6144,0.012111110819710625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,7168,0.018050667312410142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,4096,0.007297777467303806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,6144,0.016169778174824186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,5120,0.011152889165613385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,3584,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,5120,0.014389332797792224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,4096,0.010031999813185798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,3072,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,51200,0.03615644574165344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,3584,0.009509333305888707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,4096,0.012432888978057437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,2560,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,2048,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,3072,0.009158222211731805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,3584,0.011716444459226398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,1536,0.004118222329351637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,2560,0.008739555875460306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,3072,0.010653333531485664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,1024,0.003504000190231535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,2048,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,2560,0.009475555684831407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,768,0.0031537777847713898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,1536,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,2048,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,512,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,1024,0.0074720001882976955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,1536,0.007772444850868649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,256,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,1024,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,512,0.006654222392373615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,768,0.006177777631415262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,128,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,512,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,64,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,256,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,32,0.002116444417172008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,128,0.005048888838953442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,65536,0.03429066803720262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,64,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,16384,0.014896889527638754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,51200,0.027813331948386297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,65536,0.12444355752733018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,51200,0.06836266650093926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,51200,0.09798222117953831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,16384,0.012069332930776807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,768,0.007056000332037608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,12288,0.01033688916100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,16384,0.021911111142900255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,256,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,10240,0.009333333207501305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,16384,0.033732444047927856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,12288,0.01882488860024346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,128,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,12288,0.026367111338509455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,10240,0.01721955504682329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,7168,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,10240,0.023002665903833177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,6144,0.008160889148712158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,8192,0.019380443625979953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,7168,0.0131217779384719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,32,0.005092444519201915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,5120,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,7168,0.017880888448821176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,6144,0.012087110843923358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,4096,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,6144,0.016221332881185744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,5120,0.011049778097205691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,3584,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,4096,0.00999466660949919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,5120,0.01426488823360867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,3072,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,4096,0.01240888900227017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,3584,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,3584,0.011400889191362592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,3072,0.009005332986513773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,2048,0.00453333349691497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,3072,0.010454222559928894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,2560,0.008665777742862701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,2560,0.009500444763236577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,1536,0.00388711111413108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,8192,0.008422222402360704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,2048,0.00835111074977451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,2048,0.008664888640244802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,1536,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,768,0.0028497777465316984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,8192,0.014112888111008538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,1536,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,1024,0.007360888852013483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,512,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,1024,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,768,0.0069884442620807225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,256,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,512,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,65536,0.084842668639289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,768,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,128,0.002327999927931362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,256,0.006395555618736479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,64,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,128,0.006311111152172089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,256,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,32,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,128,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,2560,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,65536,0.03243733445803324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,51200,0.026075555218590632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,16384,0.011019555230935415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,64,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,12288,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,65536,0.12443999449412028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,51200,0.09774311383565266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,16384,0.03320444292492337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,1024,0.0032097777972618737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,8192,0.007831110722488826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,7168,0.007335999773608313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,10240,0.022822222775883142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,6144,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,8192,0.01926666663752662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,5120,0.007439999944633908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,7168,0.01757066614098019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,512,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,4096,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,3584,0.006224888894293044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,6144,0.015916445189052157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,3072,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,5120,0.014165333575672574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,2560,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,4096,0.012439111040698158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,2048,0.004396444393528832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,3584,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,1536,0.003882666842805015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,32,0.005095110999213325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,1024,0.0031840000301599503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,2560,0.00960622231165568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,768,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,10240,0.00870311094654931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,512,0.002536888958679305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,1536,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,256,0.002201777779393726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,12288,0.026352889007992212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,1024,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,128,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,64,0.001875555556681421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,768,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,32,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,512,0.005859555469618903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,256,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,65536,0.03028800090154012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,51200,0.025018667181332905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,128,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,16384,0.00909511082702213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,64,0.00499199993080563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,12288,0.012609778179062737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,32,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,10240,0.01034044474363327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,8192,0.00941777808798684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,65536,0.12425155109829372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,16384,0.033294220765431724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,51200,0.09787644280327691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,7168,0.008681777450773451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,12288,0.02626488937271966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,6144,0.008731555607583787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,5120,0.008357333640257517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,10240,0.022760889596409265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,4096,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,8192,0.019327110714382596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,3584,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,3072,0.010456889039940305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,3072,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,7168,0.01774666706720988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,2560,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,6144,0.01609244445959727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,2048,0.004551110996140374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,2048,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,1536,0.003573333223660787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,5120,0.014182221558358936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,1024,0.0031902222997612427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,4096,0.01240799989965227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,3584,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,3072,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,512,0.002471999989615546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,256,0.002199999988079071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,2560,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,2048,0.008586666650242275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,128,0.0018862223045693503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,1536,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,1024,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,512,0.005624889085690181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,256,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,128,0.004974222017659081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,64,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,32,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,768,0.002846222163902389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,64,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,768,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,16384,1.550564448038737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,16384,1.7925217946370442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,51200,1.8865350087483723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,16384,0.5058986875745985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,12288,1.3005848990546334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,12288,0.38378310203552246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,16384,1.5094817479451497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,12288,1.2702791425916884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,51200,4.8582712809244795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,10240,0.9894079632229276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,10240,0.33402132987976074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,10240,1.1006533304850261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,8192,0.7155155605740018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,12288,1.057976934644911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,8192,0.2711706691318088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,8192,0.8437893125745984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,51200,6.303437127007379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,10240,0.8892640007866753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,7168,0.7234400113423666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,7168,0.24318223529391816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,6144,0.6555848651462132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,8192,0.7254915767245822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,7168,0.7272871335347494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,6144,0.21270578437381318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,6144,0.6336337725321451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,5120,0.17930132812923857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,7168,0.621231132083469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,5120,0.7229653464423286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,5120,0.5310275819566515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,4096,0.3608791033426921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,6144,0.5323164198133681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,3584,0.30999554528130424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,4096,0.18512354956732857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,4096,0.4104311201307509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,5120,0.45920446183946395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,3584,0.3759937816196018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,3584,0.13303021589914957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,3072,0.3680266539255778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,4096,0.37311021486918133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,3072,0.3096906609005398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,2560,0.23424355189005533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,3584,0.3261004553900825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,3072,0.12190755208333333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,2048,0.19025244977739122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,2560,0.1050213310453627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,3072,0.28748265902201336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,2560,0.2807893223232693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,1536,0.17490222718980578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,2560,0.24115112092759874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,2048,0.22024799717797172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,1024,0.12427111466725667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,1536,0.17000088426801893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,1536,0.07068889008627997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,2048,0.20551821920606825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,768,0.08303466770384046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,1536,0.16035200489891901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,1024,0.13031466801961264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,1024,0.05668444765938652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,512,0.05765955315695869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,768,0.05092622174157036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,768,0.10254044665230645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,256,0.04088711076312595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,1024,0.13914755980173746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,51200,4.945832146538629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,512,0.04577155576811897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,512,0.08859822485182021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,128,0.037039111057917275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,768,0.10046488708919948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,256,0.05144800080193413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,256,0.07989866866005792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,512,0.0834275550312466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,128,0.04070577687687344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,32,0.03520444366667006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,128,0.0746684471766154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,2048,0.08983021974563599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,256,0.06668711370891996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,64,0.07425866524378459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,32,0.07461688915888469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,128,0.05044800043106079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,64,0.04013777772585551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,65536,1.854381349351671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,51200,1.3296737670898438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,51200,3.765079074435764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,16384,1.4720257653130426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,65536,4.866455925835504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,51200,4.880049811469184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,16384,1.5290248658921983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,16384,0.3991768889957004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,12288,0.9842728508843316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,12288,0.3026799890730116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,12288,0.9798684649997287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,65536,6.669873979356553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,10240,0.7892826928032769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,16384,1.1588684717814128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,10240,0.2626355489095052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,10240,0.8791991339789497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,12288,0.8703022003173828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,8192,0.5514728758070204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,8192,0.6873591211107043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,10240,0.688368002573649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,8192,0.2142782211303711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,7168,0.5584728982713487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,51200,3.971448686387804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,7168,0.5604497591654459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,8192,0.5627262327406142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,7168,0.19104088677300346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,6144,0.4897226757473416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,6144,0.168013334274292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,5120,0.384659555223253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,6144,0.5050942103068033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,5120,0.1421511173248291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,4096,0.31450043784247506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,5120,0.41508089171515566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,6144,0.4348453415764703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,4096,0.31254932615492076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,4096,0.12466577688852946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,3584,0.26424532466464573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,5120,0.3582151201036241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,3072,0.2118542194366455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,3584,0.28444088829888237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,3584,0.10601244370142619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,65536,5.074537065294054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,4096,0.29606933063930935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,2560,0.22499556011623809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,3072,0.09597955809699164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,3584,0.26265154944525826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,2048,0.1824400027592977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,2560,0.26114487648010254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,7168,0.4921111000908746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,3072,0.22677156660291883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,1536,0.12537422445085314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,2560,0.19455288516150582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,2048,0.0702071123652988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,2048,0.1709697776370578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,1024,0.08327822552786933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,1536,0.05662577682071262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,1536,0.13574043909708658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,2048,0.16440355777740479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,768,0.06371466981040107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,1024,0.04559288753403557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,1536,0.12694488631354436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,1024,0.1027208897802565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,512,0.05228088961707222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,768,0.04136177897453308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,1024,0.09955288966496785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,768,0.08198133442136976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,3072,0.2545155684153239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,512,0.037371555964152016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,256,0.036983999941084124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,512,0.0704560014936659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,128,0.030299554268519085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,256,0.063609778881073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,512,0.06809155808554755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,256,0.03437688946723938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,64,0.024746666351954143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,32,0.025098666548728943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,256,0.05418311225043403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,128,0.033598221010631986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,128,0.0597626699341668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,64,0.05934488773345947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,128,0.04167911079194811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,32,0.05959733327229818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,768,0.08228266901440091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,2560,0.08720799949434067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,65536,0.573943985833062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,51200,1.2785608503553603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,65536,1.703926192389594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,16384,0.42921333842807347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,51200,0.4668702019585504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,51200,1.6485075420803492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,16384,0.14753066168891057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,65536,2.1526071760389542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,12288,0.3153955671522352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,16384,0.44919464323255753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,12288,0.10941511392593384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,10240,0.3189697795444065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,12288,0.32977599567837185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,16384,0.3712755574120416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,10240,0.266957336001926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,8192,0.2535528871748182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,10240,0.10230844550662571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,8192,0.07468533515930176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,7168,0.21607555283440483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,10240,0.24026756816440156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,8192,0.1917795605129666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,7168,0.06824444399939643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,65536,1.6410800086127386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,51200,1.2451057434082031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,7168,0.19469332695007324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,6144,0.18692889478471544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,5120,0.15913244088490805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,6144,0.06946755780114068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,7168,0.17584710650973848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,6144,0.157425774468316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,5120,0.1605751117070516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,5120,0.05448444353209602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,4096,0.12296977308061387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,12288,0.28315912352667916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,6144,0.16440178288353816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,3584,0.10684355762269761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,5120,0.12634666760762533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,4096,0.10732533534367879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,4096,0.04320889049106174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,3072,0.09222222036785549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,4096,0.10822044478522407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,3584,0.03779200050565932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,2560,0.07833155658509997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,3072,0.0343466666009691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,3584,0.09495644436942206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,3072,0.08356710937288071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,2048,0.06193511353598701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,2560,0.030652443567911785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,2560,0.07211644119686551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,2048,0.0598906675974528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,2048,0.02587377859486474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,2560,0.07350755400127836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,1536,0.04489422175619337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,2048,0.0631493330001831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,1536,0.020997333857748244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,1024,0.03200799889034695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,1536,0.047543111774656505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,3584,0.09499377674526638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,1536,0.05161955621507433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,1024,0.017518222332000732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,768,0.024302222662501868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,1024,0.03640266590648227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,512,0.017272889614105225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,768,0.01628888812330034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,1024,0.04109866751564874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,768,0.030194666650560167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,256,0.01296622223324246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,512,0.01496266656451755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,3072,0.08461866776148479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,768,0.035460445615980364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,128,0.01000088867213991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,256,0.013600000076823764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,256,0.021878222624460857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,512,0.030307554536395605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,64,0.008832889298597971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,128,0.021316443880399067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,128,0.013764444324705334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,32,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,256,0.024051555328898962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,8192,0.21020800537533232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,128,0.01882133384545644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,64,0.02130488885773553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,32,0.02121777832508087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,512,0.02477688921822442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,65536,0.46099646886189777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,51200,0.9432764053344727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,51200,0.3542284435696072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,65536,1.442741288079156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,16384,0.3722657892439101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,51200,1.1607964833577473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,65536,1.4168853759765625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,12288,0.2729724513159858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,16384,0.12454488542344834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,16384,0.3341368834177653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,12288,0.10364622539944118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,10240,0.2364515463511149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,12288,0.24746221966213652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,16384,0.30243643124898273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,8192,0.19151111443837485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,12288,0.22453512085808647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,8192,0.05961866511238945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,8192,0.16571822431352404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,65536,1.248393800523546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,10240,0.2115804486804538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,7168,0.14593244923485651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,8192,0.1549039946662055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,6144,0.1258702278137207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,7168,0.0536444452073839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,7168,0.14211377832624647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,5120,0.12353066603342693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,6144,0.047041777107450694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,6144,0.1244586706161499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,7168,0.14208977752261692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,4096,0.09613511297437881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,5120,0.10389510790506999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,10240,0.0731733308898078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,5120,0.10168533192740546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,4096,0.03376977642377218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,4096,0.08504977491166855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,51200,0.9483724170260959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,4096,0.08510133292939927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,3584,0.03003022074699402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,3584,0.0739519993464152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,3072,0.07230577866236369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,3072,0.026926222774717543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,3584,0.077711111969418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,2560,0.061843554178873696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,3072,0.06497155295477973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,3072,0.0678960018687778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,2560,0.02459200057718489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,2560,0.05671199825074938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,2048,0.04715999960899353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,5120,0.0399617784553104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,10240,0.20221689012315539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,1536,0.03487199876043532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,2048,0.02120888895458645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,2560,0.05922933419545492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,2048,0.047192000680499606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,1024,0.02347022294998169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,3584,0.08615377876493667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,1536,0.017581333716710407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,1536,0.03751822312672933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,2048,0.05090399914317661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,768,0.020084444019529555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,1024,0.014866666661368476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,1024,0.02917155623435974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,1536,0.04305422306060791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,768,0.013748444616794586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,768,0.02407288882467482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,1024,0.03407288922203912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,256,0.013890667094124688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,512,0.012807110945383707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,768,0.029471112622155085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,256,0.012055111428101858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,6144,0.11899733543395996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,128,0.011501332951916588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,256,0.020693333612547982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,128,0.017099555995729234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,128,0.011924444801277585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,128,0.01679111189312405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,512,0.017044444878896076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,32,0.010685332947307162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,64,0.01721155477894677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,512,0.02000711030430264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,32,0.017250667015711468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,256,0.017656000124083627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,65536,0.9483395682440864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,64,0.010668444136778513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,65536,0.37152711550394696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,16384,0.27161423365275067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,65536,1.3068524466620552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,51200,0.9321262571546766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,51200,0.2990470992194282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,16384,0.11315733856625027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,16384,0.30867377916971844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,51200,0.918922636244032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,16384,0.2602684497833252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,12288,0.2861093415154351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,10240,0.17357777224646675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,12288,0.07271555397245619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,65536,1.063770718044705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,12288,0.19792532920837402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,8192,0.13907378249698216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,10240,0.06208266814549764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,51200,0.7954675356547037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,10240,0.16923999786376953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,7168,0.11457867092556423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,8192,0.05064622229999966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,8192,0.1369431151284112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,10240,0.16875467035505506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,6144,0.10285066896014744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,512,0.024507555696699355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,7168,0.04516355527771843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,7168,0.1192435555987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,8192,0.1380231115553114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,5120,0.09261511431799994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,6144,0.03913155529234145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,6144,0.10505066977606879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,7168,0.12093955940670437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,4096,0.07178222470813327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,5120,0.03376177615589566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,5120,0.08748622073067559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,3584,0.06540799803204007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,6144,0.10463377502229478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,4096,0.0720284448729621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,5120,0.0898959967825148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,3584,0.026692445079485577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,3072,0.052896890375349254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,4096,0.07613244321611193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,3584,0.06321066617965698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,2560,0.04284088810284933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,3072,0.023971555961502924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,3584,0.0670915577146742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,2048,0.03683999843067593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,2560,0.021739555729760066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,2560,0.0484479996893141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,12288,0.20679199695587158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,1536,0.026608000199000042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,2048,0.01917866700225406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,2560,0.053547554545932345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,2048,0.04051999913321601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,1024,0.017047999633683097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,1536,0.03291644321547614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,2048,0.04548977812131246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,4096,0.028990222348107234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,768,0.01462399959564209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,1024,0.013968000809351603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,1536,0.03753066725201077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,1024,0.025409777959187824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,512,0.01239999963177575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,768,0.012800888882742988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,768,0.02100177771515316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,1024,0.030692444907294378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,512,0.017431111799346078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,512,0.011747555600272285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,768,0.02686133318477207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,3072,0.06104089154137505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,512,0.02273066673013899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,256,0.01572355628013611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,128,0.008643555144468943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,256,0.011034666664070554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,256,0.018921777606010437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,128,0.01110311100880305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,64,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,1536,0.016632889707883198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,32,0.00795911086930169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,128,0.015159999330838522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,128,0.015861334072219003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,64,0.015152000718646579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,32,0.015103111664454142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,3072,0.055849777327643506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,256,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,65536,0.32111022207472056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,65536,0.9894373151991102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,65536,0.9506435394287109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,16384,0.27233955595228404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,51200,0.2510204315185547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,51200,0.7616400188869901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,16384,0.08852089113659328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,16384,0.22303377257453072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,12288,0.17767822742462158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,65536,0.8992150624593099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,12288,0.06893333461549547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,16384,0.2447591092851427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,51200,0.6751093334621854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,10240,0.15852444701724583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,51200,0.6380293104383681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,8192,0.12173244688245986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,12288,0.16688711113399932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,10240,0.05322133170233833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,10240,0.15092355675167507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,7168,0.09521866507000393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,8192,0.043450666798485645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,8192,0.1251537799835205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,10240,0.14168000221252441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,6144,0.0965342190530565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,7168,0.0381279985109965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,7168,0.09838577773835923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,5120,0.07248355282677545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,8192,0.11640444066789414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,6144,0.03473155697186788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,4096,0.0646373364660475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,7168,0.10362399948967828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,5120,0.029277334610621136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,5120,0.07278044356240167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,6144,0.08967110845777725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,4096,0.02516533268822564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,12288,0.1760702265633477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,5120,0.07901866568459405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,3072,0.042303999265034996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,3584,0.022992889086405437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,4096,0.06560711065928142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,3584,0.05316266748640272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,2560,0.039248890346950956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,3072,0.02126666737927331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,3072,0.047148442930645414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,3584,0.059749331739213735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,2048,0.03185955683390299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,2560,0.04074666566318936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,2560,0.019190222024917603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,3072,0.05341066585646736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,1536,0.02368800010946062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,2048,0.01713244451416863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,2560,0.046466665135489575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,6144,0.0891466670566135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,1024,0.016809778081046212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,1536,0.028055111567179363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,1536,0.014440889159838358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,2048,0.03978755407863193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,768,0.01348000019788742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,1024,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,4096,0.06031911240683662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,1024,0.021599110629823472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,512,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,1536,0.03338044550683763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,768,0.011362666885058085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,768,0.01813066667980618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,256,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,1024,0.02730488942729102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,512,0.01478488908873664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,512,0.010688888529936472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,768,0.023412444525294836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,128,0.006523555351628198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,256,0.010266666611035665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,256,0.017479999197853934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,128,0.010044444766309526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,128,0.01312711089849472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,64,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,32,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,2048,0.0344471103615231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,128,0.01367555558681488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,3584,0.0537431107627021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,64,0.0131217779384719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,32,0.01312622262371911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,65536,0.6991688940260145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,65536,0.2689173221588135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,51200,0.5593866772121853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,256,0.013457777599493662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,65536,0.8514400058322482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,16384,0.17976089318593344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,51200,0.2068533367580838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,16384,0.19838399357265898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,16384,0.06779288583331637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,51200,0.6910044352213541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,12288,0.11339910825093587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,65536,0.7374044524298774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,16384,0.18232622411515978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,12288,0.05203555689917671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,10240,0.11949243810441758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,51200,0.572439087761773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,8192,0.09757955869038899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,10240,0.048488001028696694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,10240,0.12222311231825088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,512,0.01997066703107622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,7168,0.07380177577336629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,8192,0.035975111855400935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,10240,0.12053422133127849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,8192,0.10822488864262898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,6144,0.07292444176144071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,7168,0.03269333309597439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,7168,0.08896978033913507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,5120,0.05850044223997328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,6144,0.029393778906928167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,6144,0.07818400197558932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,5120,0.02572888963752323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,5120,0.0653466648525662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,12288,0.1505591074625651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,6144,0.07893955707550049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,4096,0.0502106679810418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,3584,0.04209066761864556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,12288,0.14192710982428655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,4096,0.02197955548763275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,5120,0.06849866443210177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,3072,0.03362044360902574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,4096,0.05390755666626824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,3584,0.020234665936893888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,3584,0.048325333330366343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,2560,0.031143999762005273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,3072,0.01863911085658603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,3072,0.04292266567548116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,3584,0.05159733361668057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,8192,0.10033155812157525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,2048,0.025012445118692186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,2560,0.03734488950835334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,2560,0.0169840008020401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,3072,0.04588355620702108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,1536,0.019662222928471036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,7168,0.08740799956851536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,2048,0.014993778533405729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,1024,0.01402933398882548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,2560,0.040744000011020236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,1536,0.013425778183672162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,1536,0.02569777766863505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,768,0.011415111521879831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,1024,0.011379555695586734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,1024,0.019613333874278598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,1536,0.029583109749688044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,512,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,768,0.010602666272057427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,1024,0.024039111203617517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,768,0.016531555189026725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,256,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,4096,0.0573635564910041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,512,0.013799111048380533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,512,0.010064000056849586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,128,0.006212444355090459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,256,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,256,0.01238933371173011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,512,0.01789866718981001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,64,0.00546577779783143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,128,0.01220266686545478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,128,0.00941422250535753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,2048,0.0311119986904992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,64,0.012064889073371887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,128,0.013105777402718862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,2048,0.035135110219319664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,32,0.012076444096035428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,65536,0.2502311070760091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,65536,0.6324897872077094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,51200,0.4385537836286757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,768,0.020711110697852243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,16384,0.16899200280507407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,65536,0.7396133210923937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,32,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,51200,0.571981324089898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,16384,0.06223555405934652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,16384,0.16868178049723306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,12288,0.128675553533766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,16384,0.199325336350335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,12288,0.052301334010230176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,10240,0.11644355456034343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,51200,0.5588124593098959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,12288,0.13793688350253636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,10240,0.10963822073406643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,10240,0.042509333954917065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,8192,0.09568088584476048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,256,0.016187555260128446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,10240,0.1196168926027086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,7168,0.08472177717420791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,8192,0.08755022287368774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,8192,0.03478488988346524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,65536,0.7072177992926704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,8192,0.09718044598897298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,7168,0.07728177971310086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,12288,0.12770133548312718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,7168,0.08612089024649726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,6144,0.06770400206247966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,5120,0.058437334166632764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,6144,0.028227554427252874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,6144,0.07604622178607516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,5120,0.024930665890375774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,4096,0.046237332953347095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,51200,0.19789511627621123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,5120,0.06545422474543254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,4096,0.04759555392795139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,3584,0.040215111441082425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,4096,0.02124799953566657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,3584,0.019655999210145738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,7168,0.031488888793521456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,3584,0.04229155513975355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,3072,0.03667555583847894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,6144,0.06752622127532959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,2560,0.028904888365003798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,3584,0.04984977841377258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,3072,0.01811911165714264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,2048,0.023810666468408372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,2560,0.01628444426589542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,2560,0.032413333654403687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,2048,0.027638221780459087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,5120,0.057576000690460205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,1536,0.018743111027611625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,2048,0.014485332700941296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,1536,0.01274577776590983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,2048,0.034282667769326106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,1536,0.022687999738587275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,1536,0.02888800038231744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,4096,0.05550488829612732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,768,0.01181244436237547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,1024,0.017603556315104168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,1024,0.011131555669837527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,1024,0.023307555251651343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,512,0.008848889006508721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,768,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,3072,0.03724888960520426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,768,0.014889778362380134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,256,0.006468444648716185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,512,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,3072,0.04554399847984314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,768,0.020243555307388306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,128,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,256,0.009071110851234859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,256,0.011425777441925473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,512,0.01755199995305803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,64,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,128,0.01110222190618515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,1024,0.014527999692493014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,32,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,64,0.011061333119869232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,128,0.012792888614866467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,32,0.011058666639857821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,65536,0.45833513471815324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,65536,0.2329662111070421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,512,0.012424888710180918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,51200,0.3741004467010498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,65536,0.6267600059509277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,2560,0.03964711229006449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,16384,0.13900355497996011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,128,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,256,0.015346666177113851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,51200,0.18919911649492052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,65536,0.5634782049391005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,12288,0.10622933175828721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,16384,0.14406843980153403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,16384,0.05783910883797539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,51200,0.4846835666232639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,10240,0.09355466895633274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,12288,0.045810666349199086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,16384,0.14808177947998047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,8192,0.07675022549099393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,10240,0.03899111019240485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,12288,0.11045422156651814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,51200,0.44255910979376895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,10240,0.09243200222651164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,7168,0.06786933210160997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,12288,0.1180151171154446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,8192,0.0315324432320065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,10240,0.09824355443318684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,6144,0.04761866728464762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,8192,0.07607644134097628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,7168,0.06829955842759874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,8192,0.08096444606781006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,6144,0.05833244323730469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,7168,0.07237422466278076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,5120,0.04981244603792826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,6144,0.06433244546254475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,5120,0.022022222479184467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,4096,0.03789244426621331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,4096,0.019848000672128465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,5120,0.05486489004558987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,3584,0.0329075555006663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,4096,0.04156444470087687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,4096,0.0471937788857354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,3584,0.017605332864655387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,3072,0.02801688843303257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,3584,0.03680888811747233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,7168,0.028318223026063707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,3072,0.015779554843902588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,3072,0.032976888948016696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,5120,0.04546133346027798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,6144,0.026766222384240892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,2048,0.01937599976857503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,2560,0.028675556182861328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,2560,0.033766223324669734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,2048,0.02446222139729394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,1536,0.015118221441904703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,2048,0.013569778038395775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,2048,0.029947555727428857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,1536,0.019896000623703003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,1024,0.011123555401961008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,1536,0.011566222541862063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,1536,0.024921778175565932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,2560,0.023187556200557288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,1024,0.015440000428093804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,768,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,1024,0.010037333601050908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,2560,0.014660444524553088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,512,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,768,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,768,0.013450666434235044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,512,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,768,0.017900443739361234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,512,0.011372444530328115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,256,0.008829333715968663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,512,0.016978666186332703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,128,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,256,0.01038755559258991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,256,0.014856000741322836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,64,0.004576888763242298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,128,0.00850577818022834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,32,0.004860444615284602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,3584,0.04265066650178698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,128,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,64,0.010072000324726105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,1024,0.020399111840460036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,32,0.009974222216341231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,65536,0.49338044060601133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,256,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,65536,0.18624533547295463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,51200,0.3393662240770128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,65536,0.5151466793484157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,16384,0.11652444468604194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,51200,0.14699378278520372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,51200,0.4424604574839274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,128,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,16384,0.048828442891438804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,12288,0.0944782230589125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,65536,0.5545271237691244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,10240,0.07815288835101657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,16384,0.1455262237124973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,12288,0.040349334478378296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,51200,0.4331066608428955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,12288,0.09250311056772868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,8192,0.0618764426973131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,10240,0.034727109803093806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,12288,0.1121635569466485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,10240,0.07858310805426703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,7168,0.052384889788097806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,8192,0.027796443965699937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,8192,0.06417866547902425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,6144,0.03892799880769517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,7168,0.025127111209763422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,7168,0.05698844459321764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,8192,0.07955555783377753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,5120,0.03365866674317254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,6144,0.023041778140597876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,6144,0.04965066578653124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,16384,0.12555732991960314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,4096,0.030457778109444514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,5120,0.020482665962643094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,5120,0.04284088810284933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,6144,0.062301331096225314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,3584,0.02755822241306305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,3072,0.03808444407251146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,4096,0.03570577833387587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,5120,0.05405866768625048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,3072,0.022864888111750286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,3584,0.03212177753448486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,3584,0.016195555528004963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,4096,0.04584355486763848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,2560,0.018951111369662814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,3072,0.014818666709793938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,3072,0.028738667567571003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,3584,0.04123199979464213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,10240,0.09453688727484809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,2048,0.016753777861595154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,2560,0.013595555391576556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,2560,0.024840889705551997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,7168,0.06995555427339342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,1536,0.012598222328556908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,2048,0.020941333638297185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,2048,0.01239733315176434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,2560,0.03296355406443278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,1024,0.00904444439543618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,1536,0.011003555523024665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,2048,0.029156443145540025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,1536,0.01738133364253574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,768,0.007927999728255803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,4096,0.017720000611411203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,1024,0.013614222407341003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,1024,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,1536,0.02353511088424259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,512,0.006561777657932705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,768,0.00925333301226298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,768,0.011958222422334882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,1024,0.019581332802772522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,256,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,512,0.010232888989978367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,512,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,768,0.01760088900725047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,128,0.004613333278232151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,256,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,256,0.008336000144481659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,3072,0.0373671121067471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,128,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,128,0.0083137775460879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,32,0.004559999952713649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,128,0.011424889167149862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,64,0.008980444735950893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,32,0.008989333278603023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,65536,0.312969790564643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,65536,0.14920532703399658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,51200,0.24590844578213164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,65536,0.46020089255438906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,16384,0.09578222036361694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,51200,0.11997066603766547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,51200,0.37316534254286027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,512,0.015487111277050443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,64,0.004512888689835866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,12288,0.07288533449172974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,16384,0.11520889070298936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,16384,0.03901066713862949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,256,0.014123555686738757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,51200,0.32873066266377765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,12288,0.03194222185346816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,16384,0.11360711521572536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,12288,0.08276710907618205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,8192,0.04427199893527561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,10240,0.07042844427956475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,10240,0.027154665854242113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,12288,0.08844444486829971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,7168,0.044064889351526894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,8192,0.021903110875023737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,10240,0.07441333267423841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,8192,0.058232890235053174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,6144,0.03581600056754218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,7168,0.02032355633046892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,7168,0.051311112112469144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,8192,0.06345333655675252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,5120,0.030576000610987347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,6144,0.0184515549076928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,7168,0.05509600043296814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,6144,0.04515555500984192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,4096,0.0248355550898446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,65536,0.41885866059197324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,5120,0.016511110795868766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,6144,0.04977777931425306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,5120,0.039213332864973284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,3584,0.022823999325434368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,4096,0.014653333359294467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,10240,0.053263111246956714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,3072,0.01942044496536255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,5120,0.04348977737956577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,4096,0.0328400002585517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,2560,0.01643200053109063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,3584,0.029616001579496596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,4096,0.036887112591001726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,3072,0.012378666963842181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,2048,0.014145778285132514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,3072,0.026429333620601233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,3584,0.03415288858943515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,2560,0.011362666885058085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,1536,0.011416888899273343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,2560,0.02271911170747545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,3072,0.0310791101720598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,2048,0.010372444159454769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,2048,0.019340443942281935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,2560,0.027744000156720478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,1536,0.00941777808798684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,1536,0.0158942225906584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,2048,0.02443733314673106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,1024,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,1024,0.012683555483818054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,1536,0.020605333977275424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,768,0.011183111204041375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,768,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,512,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,3584,0.01346666696998808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,768,0.015496888094478183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,256,0.0047048889100551605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,512,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,512,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,512,0.014142221874660917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,128,0.003882666842805015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,256,0.007631111476156447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,256,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,1024,0.008719999757077958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,64,0.0035537779331207275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,128,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,256,0.012425777812798819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,128,0.008664888640244802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,32,0.00387999994887246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,768,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,128,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,64,0.008621333373917473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,65536,0.2918817732069227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,32,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,1024,0.01688088807794783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,65536,0.14266755845811632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,51200,0.22861777411566842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,65536,0.4024231168958876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,16384,0.09044355816311306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,51200,0.11254489421844482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,12288,0.06891822152667575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,16384,0.0381368895371755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,51200,0.3203608989715576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,65536,0.4104151195949978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,16384,0.09923555453618367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,10240,0.058711111545562744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,12288,0.030766222212049697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,12288,0.07563377751244439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,51200,0.3226711220211453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,16384,0.11214489407009548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,8192,0.045419553915659584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,10240,0.06480977932612102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,7168,0.0395706660217709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,8192,0.02195022172398037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,8192,0.05295022328694662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,10240,0.07270577881071302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,6144,0.03450222147835626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,7168,0.02023555503951179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,7168,0.047578665945265025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,8192,0.061533331871032715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,6144,0.018368888232443068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,7168,0.05539288785722521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,6144,0.041624890433417425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,4096,0.023787554767396715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,5120,0.016572443975342643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,5120,0.035636444886525474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,6144,0.048935112026002675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,3584,0.021544888615608215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,4096,0.03001333276430766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,4096,0.014411555396185981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,5120,0.042805333932240806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,3072,0.019124444988038804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,3584,0.013400889105266996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,3584,0.027221333649423387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,4096,0.0366960002316369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,10240,0.027215111586782668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,2560,0.0166275550921758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,12288,0.08589155806435479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,3072,0.024101333485709295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,2048,0.014356444279352823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,3584,0.03303466571701898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,2560,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,2560,0.020670221911536325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,1536,0.011744889120260874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,5120,0.02758399976624383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,2048,0.010255999863147736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,2560,0.027289777994155884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,1024,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,2048,0.023755555351575215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,1536,0.014800000521871777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,1024,0.00871022211180793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,1024,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,1536,0.019888000355826486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,768,0.007287110719415877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,768,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,1024,0.016536000702116225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,512,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,768,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,3072,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,768,0.01517511076397366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,512,0.008005333443482717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,256,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,512,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,128,0.003896000070704354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,256,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,512,0.013903111219406128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,3072,0.030249777767393324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,256,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,64,0.0038720000949170855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,2048,0.01774222155412038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,32,0.004563555535342959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,128,0.008063111040327284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,128,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,256,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,1536,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,64,0.007664889097213745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,128,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,32,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,65536,0.2902826733059353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,65536,0.1370133293999566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,51200,0.18755022684733072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,16384,0.07395378086302015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,65536,0.3572399881150987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,51200,0.11982756190829807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,12288,0.05736800034840902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,51200,0.27924709849887425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,16384,0.03681244452794393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,16384,0.08817511134677464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,65536,0.40772355927361387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,10240,0.03893333342340257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,12288,0.029792000850041706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,51200,0.3211182223425971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,12288,0.06941422488954332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,16384,0.11124177773793538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,8192,0.036976890431510076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,10240,0.026484444737434387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,10240,0.05740177631378174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,7168,0.03240888979699876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,12288,0.08562399943669637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,8192,0.021500445074505273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,6144,0.02881422307756212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,8192,0.0486195551024543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,10240,0.07378222544987996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,7168,0.020101333657900494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,5120,0.023758222659428913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,8192,0.06131111250983345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,6144,0.017667555146747164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,6144,0.0370764434337616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,4096,0.01993688941001892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,7168,0.05538488758934868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,5120,0.016015110744370353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,5120,0.032418668270111084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,6144,0.04822399881150988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,3584,0.017343110508388944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,4096,0.013953778478834363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,4096,0.027004443936877783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,5120,0.042262223031785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,3072,0.015403555499182807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,3584,0.024305777417288885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,4096,0.03628177775277032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,2560,0.013255111045307584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,3072,0.01203555530971951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,3584,0.03286222285694546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,3072,0.0218231115076277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,2048,0.01125955581665039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,2560,0.011145778000354767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,3072,0.02921244502067566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,2560,0.018933332628673978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,1536,0.009087111386987898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,2048,0.0158906661801868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,2048,0.010048888623714447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,1024,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,7168,0.042352000872294106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,2560,0.026160889201694067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,1536,0.009397333694828881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,2048,0.02370311154259576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,1536,0.013805333111021252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,768,0.005984000033802456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,1024,0.008604444563388824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,512,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,1024,0.010858666565683154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,1536,0.020071110791630216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,768,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,256,0.003931555483076307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,768,0.009663999908500248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,1024,0.01665155589580536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,768,0.0151555554734336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,3584,0.013213333156373767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,128,0.003539555602603488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,256,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,256,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,64,0.00318755561278926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,512,0.013754666679435305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,32,0.0032177778581778207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,128,0.007357333269384172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,256,0.012749333348539142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,128,0.007335111498832703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,64,0.007351111206743453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,32,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,65536,0.23119465510050455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,51200,0.1869164440366957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,65536,0.1317288875579834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,65536,0.31012354956732857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,16384,0.059327112303839795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,51200,0.2552035649617513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,51200,0.10452800326877171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,512,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,65536,0.4097324477301703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,12288,0.04640977912478977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,16384,0.034887999296188354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,51200,0.3192062113020155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,12288,0.05934488773345947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,10240,0.03885244329770406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,12288,0.028464890188641016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,128,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,12288,0.08509066369798447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,8192,0.029663112428453233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,10240,0.025399999486075506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,10240,0.05047733253902859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,8192,0.020444444484180875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,8192,0.041577776273091636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,7168,0.026132444540659588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,8192,0.06019555860095554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,16384,0.07671466800901625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,7168,0.03715822100639343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,6144,0.023342221975326538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,7168,0.01891644464598762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,7168,0.05401955710517036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,5120,0.01963644391960568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,6144,0.0329368876086341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,6144,0.017255110873116385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,5120,0.015217777755525378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,6144,0.048248890373441905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,4096,0.01685422162214915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,5120,0.02863822380701701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,4096,0.013408888545301227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,4096,0.023988443944189284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,3584,0.01424088908566369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,10240,0.07198844353357951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,4096,0.0355697770913442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,3584,0.012507555385430654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,3072,0.013252444565296173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,3584,0.021930666433440313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,512,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,2560,0.011752000285519494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,3072,0.011824000212881299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,3584,0.031481779283947416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,3072,0.01904266741540697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,16384,0.11133510536617702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,2560,0.01680088871055179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,3072,0.028724445237053767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,1536,0.008082666330867344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,2048,0.014450665977266101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,2048,0.010029333333174387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,1536,0.00920800036854214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,1536,0.012432888978057437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,1024,0.006147555592987273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,2048,0.022687999738587275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,5120,0.041723555988735624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,768,0.0053342220683892565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,1024,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,1024,0.009823111196359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,768,0.008886222210195329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,768,0.008050666915045844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,1024,0.01608800060219235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,2048,0.009739555418491364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,768,0.015073777900801765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,256,0.0038648889296584656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,512,0.00795999997191959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,2560,0.01074488874938753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,512,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,128,0.003216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,256,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,2560,0.025117332736651104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,256,0.007196444604131911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,64,0.0035386664999855887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,128,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,128,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,32,0.0038568890757030914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,256,0.012144889268610211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,64,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,1536,0.018650665879249573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,128,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,32,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,51200,0.15241066614786783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,512,0.004552888787455029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,65536,0.0875564416249593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,65536,0.2591128879123264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,16384,0.04895377821392483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,51200,0.06944711340798272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,51200,0.2039653327729967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,12288,0.034359110726250544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,65536,0.26821332507663304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,16384,0.06621866756015353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,16384,0.02462933295302921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,512,0.013423999978436364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,51200,0.2118755578994751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,12288,0.020626667473051283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,16384,0.07478577560848661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,8192,0.023621333969963923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,10240,0.018562666243977018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,10240,0.04301866557863024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,12288,0.059635553095075816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,7168,0.02095555596881443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,65536,0.19208266999986437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,10240,0.05156977640257942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,8192,0.014817777607176038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,8192,0.036010666026009455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,6144,0.018581334087583754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,7168,0.013801777528391944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,7168,0.03273244367705451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,8192,0.04364711046218872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,5120,0.01609511176745097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,6144,0.02879022227393256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,6144,0.012808889150619507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,7168,0.03871466716130575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,4096,0.013702222042613559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,5120,0.011743110915025076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,5120,0.02495288848876953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,6144,0.03494933247566223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,3584,0.012320888539155325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,10240,0.030262221892674763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,4096,0.0207022229830424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,5120,0.030304888884226482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,4096,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,12288,0.050867554214265615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,3584,0.010008888940016428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,3584,0.018936889039145578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,4096,0.026038222842746313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,2560,0.010000000397364298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,3072,0.00942844483587477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,3072,0.01690933273898231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,3584,0.02374222212367588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,2048,0.008364444805516137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,2560,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,3072,0.02199466692076789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,2560,0.014863999353514777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,1536,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,2048,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,2048,0.012905778156386482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,2560,0.01960711181163788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,1024,0.005184000151024925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,1536,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,1536,0.00813688917292489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,768,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,1024,0.009039999710188972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,1536,0.01517688896920946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,512,0.003887999802827835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,1024,0.013437333206335703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,768,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,768,0.008240888516108194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,256,0.0032035555276605818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,512,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,512,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,768,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,128,0.0028888889484935333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,256,0.006665777828958299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,256,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,512,0.012087110843923358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,64,0.0028497777465316984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,128,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,32,0.0029057777590221832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,128,0.006545777950021956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,65536,0.12952089309692383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,64,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,128,0.008625778059164682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,2048,0.017673777209387887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,32,0.006312888943486744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,51200,0.09647199842664932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,1024,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,65536,0.0842204425070021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,16384,0.03792622354295518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,65536,0.21124711301591662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,51200,0.06908000177807279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,3072,0.01089244418674045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,51200,0.16656088829040527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,12288,0.026537777649031744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,65536,0.2639262146419949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,16384,0.05506044295099047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,10240,0.0225591113169988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,51200,0.20852088928222656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,12288,0.04236177934540642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,12288,0.019648889700571697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,8192,0.018208889497650992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,16384,0.07409955395592584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,12288,0.05769689215554131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,256,0.011017777853541903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,10240,0.017847999930381775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,10240,0.03645777702331543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,7168,0.01622577839427524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,8192,0.01425066590309143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,8192,0.03036266565322876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,7168,0.013416888813177744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,6144,0.014700444208251106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,10240,0.04973155591222975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,7168,0.027285334136750963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,5120,0.01259022206068039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,6144,0.012404444317022959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,8192,0.0420702232254876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,4096,0.010620444185203977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,5120,0.011424000064531961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,5120,0.020793777373101976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,6144,0.03380266825358073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,3584,0.009937778115272522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,5120,0.029134223858515423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,4096,0.010138666464222802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,4096,0.01756533318095737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,16384,0.023464888334274292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,3072,0.008665777742862701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,3584,0.009672000176376766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,4096,0.02488888965712653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,3584,0.01609422266483307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,2560,0.007728888756699032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,3072,0.01455200049612257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,3072,0.009322666459613377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,3584,0.022786666949590046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,2048,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,2560,0.009014222357008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,2560,0.012784889174832238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,3072,0.021314667330847845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,1536,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,2048,0.008631111019187504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,2560,0.018981332580248516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,7168,0.03763555486996969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,1024,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,1536,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,1536,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,2048,0.01720888912677765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,768,0.0040737779604064096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,1024,0.007632888853549957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,1536,0.015048000547620984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,1024,0.008018666671382057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,512,0.0035217776894569397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,768,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,768,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,1024,0.013109332985348172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,256,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,512,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,512,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,768,0.012075555821259817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,128,0.002847111059559716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,256,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,256,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,64,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,512,0.011077333655622272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,32,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,128,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,128,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,2048,0.011423110961914062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,256,0.010612444745169746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,65536,0.10821511348088582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,6144,0.02381244467364417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,64,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,128,0.008617777791288164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,32,0.005969777703285217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,51200,0.08443733056386311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,16384,0.031127111779318914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,65536,0.08454577790366279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,65536,0.17281156116061738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,51200,0.06815822256935967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,12288,0.023545778459972803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,16384,0.02314311100376977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,51200,0.13616000281439886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,16384,0.046126223272747464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,10240,0.019966221517986722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,12288,0.01927911076280806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,51200,0.19900088840060762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,12288,0.03535111082924737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,8192,0.016501333978441026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,10240,0.01756533318095737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,10240,0.03088266650835673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,7168,0.014832889040311178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,12288,0.05556622478697035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,10240,0.047749333911471896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,8192,0.02551199992497762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,6144,0.013427555561065674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,7168,0.023137778043746948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,8192,0.04035022192531162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,7168,0.03625422385003831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,6144,0.012087110843923358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,6144,0.020247111717859905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,6144,0.03154222170511881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,5120,0.017989334132936265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,4096,0.010025777750545079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,65536,0.2519377867380778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,5120,0.027906666199366253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,16384,0.07006311416625977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,3584,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,4096,0.015127110812399121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,4096,0.010017777482668558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,8192,0.014274666706720987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,3072,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,3584,0.013857777747843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,3584,0.009679999616410997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,4096,0.024140444066789415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,2560,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,7168,0.013129777378506131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,3072,0.009372444616423713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,3584,0.022050667140218947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,3072,0.012732444538010491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,2048,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,5120,0.011622221933470832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,2560,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,2560,0.008952889177534316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,1536,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,3072,0.02045866681469811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,2048,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,2048,0.010116444693671333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,1024,0.0041831110914548235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,1536,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,1536,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,2048,0.016572443975342643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,5120,0.011377777490350934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,768,0.0036640001667870414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,1024,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,1536,0.014486221803559197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,1024,0.0075911109646161394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,512,0.0031786666562159858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,768,0.007439110842016008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,768,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,1024,0.012789333032237159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,256,0.002895111011134254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,512,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,512,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,128,0.0028124445428450904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,512,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,256,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,128,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,256,0.01072799993885888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,128,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,64,0.0025351111673646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,32,0.002504000026318762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,65536,0.06986755794949003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,2560,0.018551111221313477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,128,0.008404444489214156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,64,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,32,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,51200,0.060954663488599986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,65536,0.08480978012084961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,65536,0.16554666890038386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,16384,0.022805333137512207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,51200,0.1302782164679633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,51200,0.06816889180077447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,65536,0.248672882715861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,768,0.012441777520709567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,12288,0.01906844476858775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,16384,0.022522666388087805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,51200,0.19747911559210884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,256,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,16384,0.044237332211600415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,10240,0.017229333519935608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,12288,0.018937778141763475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,8192,0.013842666314707862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,12288,0.034677333301968045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,10240,0.017228444417317707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,10240,0.02962755494647556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,7168,0.012366222010718452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,12288,0.05475377705362108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,8192,0.013809777796268463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,8192,0.02442488902144962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,10240,0.047203554047478564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,6144,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,7168,0.013038222160604266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,5120,0.009648888475365108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,7168,0.022348443667093914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,8192,0.039501332574420504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,6144,0.012069332930776807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,4096,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,6144,0.019807111885812547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,7168,0.0353724459807078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,5120,0.01106844428512785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,3584,0.007191999918884701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,5120,0.0173973325226042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,6144,0.03135644396146139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,4096,0.009993777506881291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,4096,0.014900444282425774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,5120,0.027248889207839966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,3584,0.0096133334769143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,3584,0.013759999639458127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,2560,0.0058151111006736755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,4096,0.024077332682079736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,3072,0.009189333352777693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,16384,0.07045955790413751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,3072,0.012441777520709567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,2048,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,2560,0.0086666668454806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,2560,0.01108977778090371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,3072,0.01982666717635261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,1536,0.004561777744028303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,2048,0.009931555224789513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,2048,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,2560,0.018233777748213876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,2048,0.016531555189026725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,1536,0.008648888932334052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,1536,0.014124444789356656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,1024,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,768,0.0031555555760860443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,1024,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,3072,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,768,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,1024,0.012446222205956778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,768,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,512,0.00685777763525645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,768,0.011776888536082374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,256,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,512,0.006230222268237009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,512,0.010748444332016839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,256,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,128,0.0023644444429212147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,256,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,1024,0.003536888708670934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,256,0.009866666462686326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,1536,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,64,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,128,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,128,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,65536,0.04478133387035794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,128,0.007979555262459649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,64,0.005347555710209741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,32,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,51200,0.03597066799799601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,65536,0.12503733899858263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,3584,0.02184088859293196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,51200,0.09848888715108235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,65536,0.24660444259643555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,16384,0.01388533330625958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,51200,0.0676764448483785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,51200,0.19550755288865831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,16384,0.021712000171343487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,16384,0.034168889125188194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,32,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,16384,0.06876533561282687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,12288,0.01862133377128177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,10240,0.01220533334546619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,12288,0.026947556270493403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,512,0.0028640000770489373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,12288,0.05369155605634054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,10240,0.01719288859102461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,8192,0.010796444283591377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,8192,0.013815110756291283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,10240,0.046210666497548424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,7168,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,8192,0.03934311204486423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,7168,0.01792977750301361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,7168,0.012753778033786349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,6144,0.010780444575680627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,12288,0.013398222625255585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,7168,0.0342497792508867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,5120,0.008649778034951951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,6144,0.011886222494973077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,6144,0.01616533266173469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,4096,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,5120,0.011042666931947073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,5120,0.014446222119861178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,6144,0.03049600124359131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,3584,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,10240,0.02334311107794444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,5120,0.027261333333121404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,4096,0.010016889207892949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,4096,0.012447111308574677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,3072,0.005956444475385878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,3584,0.00941955546538035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,8192,0.01959822244114346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,3584,0.011615999870830111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,2560,0.005542222410440445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,4096,0.02346222268210517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,3072,0.00906844437122345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,2048,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,2560,0.00865955568022198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,3072,0.019966221517986722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,2560,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,1536,0.0038559999730851916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,2048,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,2048,0.00868622213602066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,1536,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,2048,0.016195555528004963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,1024,0.003301333429084884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,1536,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,1024,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,65536,0.08334489001168145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,1536,0.01407911048995124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,768,0.0031386667655573953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,1024,0.006648889018429651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,512,0.00282222218811512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,768,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,1024,0.012421333127551608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,768,0.006312888943486744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,256,0.002528888897763358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,3072,0.010618666807810465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,512,0.00573333352804184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,512,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,768,0.01147822207874722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,128,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,256,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,256,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,512,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,64,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,32,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,2560,0.01789955629242791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,128,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,65536,0.033964445193608604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,256,0.01034400032626258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,128,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,51200,0.027669333749347266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,32,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,65536,0.08382488621605767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,65536,0.1248133314980401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,16384,0.012063110868136087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,51200,0.06739199823803373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,51200,0.09814400143093532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,65536,0.24687022633022734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,16384,0.03384000062942505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,12288,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,51200,0.1954737769232856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,16384,0.02164622147878011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,3584,0.02130399975511763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,10240,0.008847999903890822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,16384,0.06886577606201172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,12288,0.018583110637134977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,12288,0.02648799949222141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,8192,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,128,0.006343110981914733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,10240,0.01702311138312022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,10240,0.022861333356963262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,7168,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,12288,0.05243200063705444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,64,0.005130666825506422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,8192,0.013794666363133324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,8192,0.01939733326435089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,6144,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,10240,0.04602311054865519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,7168,0.01273333364062839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,8192,0.03813955518934462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,5120,0.006670222100284364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,6144,0.011758222348160215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,6144,0.01593244406912062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,7168,0.033950222863091364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,4096,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,5120,0.010967111421955956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,6144,0.0303422212600708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,5120,0.014120888378885059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,3584,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,4096,0.00999466660949919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,4096,0.012465777496496836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,5120,0.026935110489527386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,3584,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,4096,0.023366222778956097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,3072,0.009027555584907532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,3584,0.021320000290870667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,2560,0.005032000028424793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,3072,0.01994311147265964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,2560,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,2048,0.004552888787455029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,2560,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,2560,0.017889777819315594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,1536,0.0038524443904558816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,2048,0.008309333688682979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,2048,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,1536,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,1536,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,3072,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,3584,0.011671110987663269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,1536,0.014151111245155334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,768,0.002879111096262932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,1024,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,1024,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,3072,0.010558221903112201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,512,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,1024,0.012251555919647217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,768,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,768,0.006134222365087933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,256,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,7168,0.01774577796459198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,512,0.0058666666348775225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,256,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,256,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,512,0.01071200023094813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,256,0.009725333087974125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,2048,0.016180444094869826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,1024,0.0032053333189752367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,128,0.005193777796294955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,32,0.002190222135848469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,128,0.0063528890411059065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,128,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,65536,0.03207822309599982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,51200,0.028819554381900366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,64,0.004990222139490975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,16384,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,32,0.005064888960785336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,12288,0.009688888986905416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,65536,0.12444710731506348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,51200,0.0976417793167962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,512,0.006701333241330252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,16384,0.03315644462903341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,8192,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,768,0.011359111302428775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,7168,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,12288,0.026340444882710774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,6144,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,5120,0.006996444529957241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,10240,0.022689777943823073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,4096,0.00657066661450598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,8192,0.019278221660190158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,7168,0.017602667212486267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,6144,0.015831111205948722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,3072,0.005535999933878581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,5120,0.014159111513031853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,2560,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,64,0.0019128889673286015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,4096,0.012240889171759287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,2048,0.004505777938498391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,3584,0.011424889167149862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,1536,0.003818666769398583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,2560,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,1024,0.0031928889867332247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,2048,0.008623999853928884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,768,0.00282222218811512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,1536,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,512,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,10240,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,1024,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,256,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,128,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,768,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,64,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,512,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,128,0.002159111201763153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,256,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,128,0.005039999882380168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,51200,0.024466666910383437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,3584,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,64,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,16384,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,32,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,12288,0.011722666521867117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,10240,0.010454222559928894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,65536,0.12374311023288304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,16384,0.03293244375122918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,8192,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,7168,0.00850844466023975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,12288,0.026144888665941026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,3072,0.010500444306267632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,10240,0.02274311085542043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,5120,0.0069848886794514125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,8192,0.019292443990707397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,4096,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,7168,0.017602667212486267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,3584,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,6144,0.01588622232278188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,3072,0.00553688903649648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,5120,0.014110222458839417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,2560,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,4096,0.012378666963842181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,2048,0.004423999951945411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,3584,0.01143200033240848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,1536,0.003584889074166616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,65536,0.03388888968361749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,3072,0.010398222340477837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,1024,0.002887111157178879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,768,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,2048,0.008614222208658854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,512,0.0024622221373849446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,1024,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,256,0.0021226666867733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,51200,0.09785688585705227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,128,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,512,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,6144,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,64,0.001806222316291597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,32,0.00179822225537565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,256,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,128,0.005024888863166173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,64,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,32,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,2560,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,1536,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,16384,1.2970995373196073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,16384,0.49184709125094944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,12288,0.9073919720119901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,16384,1.5965617497762044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,12288,1.067855093214247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,12288,0.37834933069017196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,10240,0.8839848836263021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,10240,0.32255376709832084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,51200,1.6106355455186632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,8192,0.5845573213365343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,10240,0.9145573510064019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,8192,0.2595751020643446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,7168,0.5860408676995171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,8192,0.7277155452304416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,6144,0.43348177274068195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,7168,0.2304933336046007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,6144,0.19941688908471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,5120,0.3700151178571913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,5120,0.17235733403099907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,6144,0.5473057958814833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,51200,3.9872762891981335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,7168,0.7901155683729383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,4096,0.3419671058654785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,5120,0.44705512788560653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,4096,0.14306488302018908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,3584,0.27526312404208714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,3584,0.12433511681026882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,4096,0.36162487665812176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,3072,0.2814115683237712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,2560,0.19745422734154594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,3072,0.1127671135796441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,2048,0.18210221661461723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,2560,0.09758400254779392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,3584,0.3052373462253147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,2560,0.22335555818345812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,2048,0.08248533142937554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,2048,0.18256799379984537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,1024,0.10330933332443237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,1536,0.064538664287991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,51200,5.101908365885417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,768,0.07664799690246582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,1536,0.14376533031463623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,1024,0.05138755506939358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,1024,0.10850489139556885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,768,0.0467564430501726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,256,0.03931466738382975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,768,0.08872799740897284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,512,0.042819556262758046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,128,0.029133333100212946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,512,0.07572533024681939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,256,0.03948533203866746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,64,0.02544622288809882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,256,0.06864266925387912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,128,0.03853866789076064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,3072,0.2625928984747993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,32,0.026221333278550044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,128,0.06413688924577501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,1536,0.13325866063435873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,64,0.06393244531419542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,32,0.06406755579842462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,512,0.05665689044528537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,16384,0.9551768832736545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,16384,0.3848115603129069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,16384,1.1592071321275499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,51200,1.2482186423407662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,65536,1.7805858188205297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,51200,3.091634750366211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,12288,0.3481137752532959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,12288,0.8438764678107368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,12288,0.8274017969767252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,10240,0.6886871125963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,10240,0.25264178382025826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,8192,0.545563538869222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,65536,3.9606124030219187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,51200,3.9033415052625866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,8192,0.5390008820427789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,8192,0.1995911068386502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,7168,0.5209182103474935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,10240,0.7097911304897733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,6144,0.3903368843926324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,6144,0.15501599841647679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,7168,0.48703824149237734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,7168,0.1790426704618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,5120,0.4089031219482422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,5120,0.3420364326900906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,6144,0.4088933202955458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,4096,0.2856968773735894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,5120,0.13622844219207764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,4096,0.11074311203426784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,3584,0.2431591086917453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,65536,5.234146542019315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,2560,0.1767022212346395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,3072,0.22859732309977213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,3072,0.08685866991678874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,3584,0.23550576633877227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,3072,0.2118533319897122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,2048,0.15714666578504774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,2560,0.0770933363172743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,2560,0.18130400445726183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,2048,0.14297866821289062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,1024,0.07434578074349298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,1536,0.052024887667761914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,2048,0.06589688857396443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,768,0.06447022491031222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,1536,0.11311733722686768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,1024,0.040820442967944674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,512,0.0529368884033627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,1024,0.08629955848058064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,768,0.03815911213556925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,768,0.0695306658744812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,256,0.03542844454447428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,512,0.03600088755289713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,512,0.06052177482181125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,256,0.03249333302179972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,4096,0.27425066630045575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,64,0.020270221763186984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,3584,0.10113244586520725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,32,0.020727111233605277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,128,0.031750222047170006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,128,0.05202933483653598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,64,0.05117511086993747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,32,0.051256000995635986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,1536,0.10806400246090359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,128,0.023390221926901076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,65536,0.5556613074408637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,256,0.055233776569366455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,51200,1.069262186686198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,65536,1.4013715320163305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,16384,0.3251013225979275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,51200,0.4386044343312581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,16384,0.13641599814097086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,12288,0.27705245547824436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,12288,0.1015591091579861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,10240,0.2658533255259196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,16384,0.430784887737698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,51200,1.205092430114746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,65536,1.5955751207139757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,12288,0.27606222364637584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,8192,0.22712710168626574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,10240,0.08652533425225152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,8192,0.07222400108973186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,7168,0.1813582181930542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,8192,0.18365600374009874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,6144,0.17043377293480766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,7168,0.07987022399902344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,6144,0.05630577935112847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,5120,0.11620799700419109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,7168,0.1577191087934706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,6144,0.13653777705298528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,4096,0.11668977472517227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,5120,0.06180533435609606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,4096,0.039679110050201416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,3584,0.10475022262997097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,5120,0.12149243884616429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,3072,0.08367288774914211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,4096,0.09352711174223159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,3584,0.03888355692227682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,2560,0.058232002788119845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,3584,0.08290133211347792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,3072,0.03289511137538486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,2048,0.04678755667474535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,3072,0.0718088878525628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,1536,0.03715555535422431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,2560,0.031423999203575984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,1024,0.02805244591501024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,2048,0.024710221423043147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,2560,0.06210933128992716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,1536,0.020219556159443326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,768,0.021944888763957556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,2048,0.05158399873309665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,512,0.016182222300105624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,1024,0.01721244388156467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,1024,0.03153422143724229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,768,0.015798222687509324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,256,0.011898666620254517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,10240,0.23881599638197157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,512,0.014512000812424554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,768,0.02656800051530202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,128,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,512,0.022086222966512043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,256,0.013198222551080914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,128,0.013144888811641268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,64,0.007771555748250749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,32,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,128,0.018623999423450895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,64,0.018938667244381376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,32,0.018602665927675035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,1536,0.0410755541589525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,65536,0.4164915614657932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,51200,0.6902302106221517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,16384,0.2657528983222114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,256,0.01927911076280806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,16384,0.11611999405754937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,65536,1.2674871020846896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,51200,0.9403582678900825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,16384,0.29001421398586696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,10240,0.17098666561974418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,10240,0.06837244166268243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,12288,0.20112621784210205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,8192,0.12512266635894775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,10240,0.16975643899705675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,7168,0.12509066528744167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,8192,0.055974225203196205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,8192,0.13786310619778103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,6144,0.10763021972444321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,7168,0.054775999652014844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,65536,0.9410141838921441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,7168,0.12132978439331055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,5120,0.0928293334113227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,6144,0.043453332450654775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,51200,0.33576001061333555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,4096,0.08284444279140897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,5120,0.03791022300720215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,6144,0.10549422105153401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,3584,0.07413688633177015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,4096,0.03109244505564372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,3072,0.05320177806748284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,12288,0.17138666576809355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,5120,0.08837244245741104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,3584,0.028703110085593328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,4096,0.07271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,2560,0.043523556656307645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,2048,0.04178933302561442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,3072,0.026309332913822595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,12288,0.08049333095550537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,3584,0.06416889031728108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,1536,0.029927998781204224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,2560,0.023242667317390442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,3072,0.056290666262308754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,1024,0.021010667085647583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,2048,0.020218667056825426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,2560,0.04861688945028517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,768,0.01626133256488376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,1536,0.017098666893111337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,2048,0.04027022255791558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,512,0.012644444074895648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,1024,0.014474666780895658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,1536,0.033033778270085655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,256,0.009249777429633671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,1024,0.025578666064474318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,512,0.012320000264379712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,128,0.007078222102589077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,768,0.02134311033619775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,64,0.0063279999627007385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,256,0.011357333097192975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,32,0.006568000134494569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,128,0.01144266708029641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,256,0.015655999382336933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,128,0.015331554743978711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,64,0.015459555718633862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,32,0.01517777807182736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,65536,0.3607164488898383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,65536,0.7791422208150228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,51200,0.5976870854695638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,768,0.01329422245422999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,16384,0.2209493319193522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,512,0.017639999588330586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,65536,1.0000933541191948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,51200,0.8117929034762912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,16384,0.08894311057196723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,12288,0.16524267196655273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,51200,0.2892675664689806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,16384,0.2438382307688395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,12288,0.06845600075191922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,8192,0.13910310798221165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,12288,0.17015555169847277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,10240,0.06353422005971272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,7168,0.12222222487131755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,8192,0.04765689041879442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,10240,0.14242666297488743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,6144,0.10798577467600505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,8192,0.11828266249762641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,7168,0.04296977652443779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,6144,0.03702311052216424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,5120,0.0951217810312907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,7168,0.10249688890245225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,4096,0.07343022028605144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,6144,0.09034755494859482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,5120,0.03335377905103896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,3584,0.061885330412122935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,4096,0.027472888429959614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,3072,0.05292710993025038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,5120,0.07536088758044772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,4096,0.07020266850789388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,2560,0.038881778717041016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,3584,0.025251555773946974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,2048,0.035790221558676824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,3584,0.05537955628501045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,1536,0.027447111076778833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,2560,0.020503999458418954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,3072,0.04828888840145535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,2048,0.018449778358141582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,1024,0.0188737776544359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,2560,0.047994666629367404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,2048,0.03536533315976461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,768,0.014874666929244995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,10240,0.17834666040208605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,1536,0.015769778026474845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,512,0.011267555256684622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,1024,0.013189333180586496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,1536,0.028939555088678997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,256,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,768,0.012102222277058495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,1024,0.022258667482270136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,128,0.006464889066086874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,512,0.011297778122954898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,64,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,768,0.01885511146651374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,32,0.005759999983840519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,256,0.010688888529936472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,512,0.01574222246805827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,128,0.010695111420419482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,256,0.013905778527259827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,128,0.013743999931547375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,64,0.013463111387358772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,32,0.013784888717863293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,16384,0.2070008913675944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,3072,0.02274488906065623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,65536,0.3069937758975559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,51200,0.5241991149054633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,65536,0.6819439993964301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,12288,0.1460204389360216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,16384,0.07460088862313165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,65536,0.7913671069675021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,51200,0.6248017946879069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,10240,0.11480889055464004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,16384,0.18733066982693145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,12288,0.06849866443210177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,12288,0.15217955907185873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,7168,0.08824533224105835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,8192,0.09687466753853692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,10240,0.05015200045373705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,8192,0.040864888164732195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,10240,0.11748711268107097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,6144,0.07734400033950806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,8192,0.09656177626715766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,6144,0.03198755449718899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,7168,0.08632622162501018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,4096,0.04405866728888618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,6144,0.07430755429797702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,5120,0.027999109692043726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,5120,0.0641440020667182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,3584,0.04541333185301887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,4096,0.051270223326153226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,51200,0.2462551063961453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,3584,0.021783111823929682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,3584,0.04520000020662943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,2560,0.03084889054298401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,3072,0.01997777819633484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,3072,0.040306667486826576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,2560,0.018200889229774475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,2048,0.025455110602908667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,2560,0.03525333272086249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,1536,0.020063110523753695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,2048,0.016192888220151264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,7168,0.04310933417744107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,1024,0.015452444553375244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,2048,0.02971733278698391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,5120,0.05264622304174635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,768,0.012891555825869242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,1536,0.02427822185887231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,1024,0.011694221860832639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,512,0.010020444790522257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,1024,0.018582221534517076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,768,0.01108088923825158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,4096,0.02411199940575494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,256,0.007536888950400882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,512,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,3072,0.03393244412210252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,512,0.013450666434235044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,128,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,256,0.009647111097971598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,64,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,256,0.012143111063374413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,32,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,128,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,128,0.011730666789743634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,64,0.011757333245542316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,32,0.011759999725553723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,65536,0.5475840038723415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,51200,0.3807226551903619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,65536,0.251888010236952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,16384,0.1442426708009508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,12288,0.09298755725224812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,16384,0.06805600060356988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,768,0.016000888413853116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,51200,0.20296266343858507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,16384,0.17129688792758516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,10240,0.09456000063154434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,12288,0.051401777399910815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,8192,0.07533511188295153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,12288,0.12524355782402888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,10240,0.042881776889165245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,7168,0.07182400094138251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,8192,0.03421511252721151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,8192,0.08842400047514175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,6144,0.05322133170233833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,7168,0.0313022236029307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,7168,0.0758497781223721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,5120,0.051500442955229014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,6144,0.02759733299414317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,4096,0.03619733452796936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,1536,0.013473778135246701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,6144,0.0664880010816786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,5120,0.024745777249336243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,3584,0.037046220567491316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,5120,0.05611911084916857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,4096,0.021075555019908484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,3072,0.03189244535234239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,3584,0.019967110620604623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,65536,0.699485354953342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,3584,0.04097244474622939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,2560,0.02646133303642273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,51200,0.5570586522420248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,3072,0.018050667312410142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,2048,0.021988444858127173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,1536,0.017071111334694754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,3072,0.03639199998643663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,2560,0.01646133263905843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,1024,0.0120328888297081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,2048,0.014498665928840637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,2560,0.03163911236657037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,2048,0.027066666218969557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,768,0.009718221922715506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,10240,0.1143146620856391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,1536,0.02237422267595927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,512,0.007557333343558841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,1024,0.017099555995729234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,256,0.005710222240951326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,512,0.009404444032245213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,768,0.014661333627170987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,512,0.012440889245933957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,128,0.005396444350481033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,256,0.009026666482289633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,256,0.011424000064531961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,64,0.004849777867396672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,128,0.009211555951171452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,32,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,128,0.011065777805116443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,4096,0.045588443676630654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,64,0.011028444601429833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,32,0.011041777829329172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,1536,0.012680889003806643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,65536,0.4890009032355414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,1024,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,65536,0.2363342179192437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,768,0.010129777921570672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,51200,0.38448532422383624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,16384,0.12166044447157119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,12288,0.10062933630413479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,65536,0.612303098042806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,51200,0.1935875548256768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,10240,0.08885511424806382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,16384,0.14827733569675022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,12288,0.056723554929097496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,12288,0.1095235546429952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,51200,0.4854213396708171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,8192,0.07549955447514851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,7168,0.0690444442960951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,10240,0.042949332131279834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,8192,0.0335822237862481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,10240,0.09601155254575942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,6144,0.0573199987411499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,8192,0.07932621902889676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,7168,0.030530667967266504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,5120,0.041109333435694374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,7168,0.06801955567465888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,6144,0.0276826669772466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,4096,0.038281778494517006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,3584,0.0340666671593984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,6144,0.05848889218436348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,4096,0.021363554729355708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,5120,0.050512890021006264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,4096,0.041675554381476514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,3584,0.03754577702946133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,2560,0.023665777511066858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,3072,0.017480888300471835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,3072,0.03313510947757297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,16384,0.059824890560574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,2048,0.020629333125220407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,2560,0.015537778536478678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,1536,0.015818667080667283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,2560,0.029241777128643457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,2048,0.014039110806253223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,1024,0.011328000161382886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,2048,0.0244862222009235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,768,0.009400000174840292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,1536,0.020243555307388306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,1024,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,1024,0.01551822159025404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,512,0.007039111521508958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,768,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,256,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,5120,0.023944889505704243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,768,0.013423999978436364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,512,0.009347555538018545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,128,0.004867555366622077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,3072,0.029110223054885864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,512,0.011438222395049201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,64,0.00423555572827657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,256,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,32,0.004599110947714912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,128,0.008959111240175035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,256,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,128,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,64,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,32,0.010011555420027839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,65536,0.40017422040303546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,65536,0.22072888745201957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,1536,0.012466666599114736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,51200,0.3795271184709337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,51200,0.17508978313869902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,16384,0.12375556098090278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,12288,0.09488266706466675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,65536,0.5224080085754395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,16384,0.05520177880922953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,10240,0.08603288729985555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,8192,0.06580177942911784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,12288,0.045888000064426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,16384,0.1252853340572781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,10240,0.03828444414668613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,12288,0.0950826671388414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,7168,0.05770844221115112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,6144,0.046365333928002246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,10240,0.08192444509930082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,3584,0.019144000278578866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,5120,0.04044533438152737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,7168,0.027687110834651526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,8192,0.07103200091256036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,6144,0.02558044426971012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,7168,0.05912355581919352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,6144,0.051923556460274585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,5120,0.021799999806616042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,3584,0.028939555088678997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,5120,0.04422488808631897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,4096,0.019184889064894784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,4096,0.036943111154768206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,3072,0.024684444069862366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,3584,0.017014222012625802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,2560,0.019938665959570143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,3584,0.032919999625947736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,3072,0.015604444675975375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,2048,0.017356443736288283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,3072,0.029848890172110662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,51200,0.3993564446767171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,2560,0.013915555344687568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,1536,0.013647110925780402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,2560,0.02608799934387207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,1024,0.009752000371615091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,2048,0.012811555630630918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,2048,0.022156443860795762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,1536,0.010989333192507425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,768,0.008039111064540016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,8192,0.031249778138266668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,1024,0.009648888475365108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,1536,0.018051554759343464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,512,0.006400888992680444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,256,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,1024,0.013815999858909182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,768,0.009014222357008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,128,0.0048177776237328844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,512,0.008647110727098253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,768,0.012340444657537671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,64,0.004233777936961916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,256,0.00848711116446389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,512,0.010464888479974536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,32,0.0041742221348815495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,256,0.00944622192117903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,128,0.008237332933478886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,128,0.00924355536699295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,64,0.009333333207501305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,32,0.009312888814343346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,65536,0.334187560611301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,65536,0.17725600136650932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,51200,0.26625333891974556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,16384,0.09610488679673936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,51200,0.1396284500757853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,12288,0.07340177562501696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,16384,0.0473235547542572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,51200,0.33062754737006295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,10240,0.061808890766567655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,16384,0.10471733411153157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,12288,0.038478222158220075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,8192,0.05173333485921224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,12288,0.0809111131562127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,10240,0.03260622090763516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,4096,0.031242665317323472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,7168,0.04131111171510484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,10240,0.067994667424096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,8192,0.026555554734336004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,6144,0.038591110044055514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,8192,0.05659111340840658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,5120,0.03247022297647264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,7168,0.02459200057718489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,4096,0.025567111041810777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,6144,0.022055110997623865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,7168,0.05029333300060696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,5120,0.019691555036438834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,3584,0.023482667075263128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,6144,0.0436115562915802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,3072,0.020451555649439495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,4096,0.016876444220542908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,5120,0.037567999627855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,2560,0.017993777990341187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,3584,0.01573866605758667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,4096,0.03149155444569058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,2048,0.01521511044767168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,3072,0.014493332968817817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,3584,0.02853333287768894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,1536,0.012489777472284106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,2560,0.013108444710572561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,1024,0.009119999905427298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,2560,0.022343110707071092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,2048,0.011936888926559024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,768,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,1536,0.01035733355416192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,2048,0.018982221682866413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,512,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,1536,0.015610666738616096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,1024,0.009335999687512716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,256,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,768,0.008804444637563493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,1024,0.012329777909649743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,512,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,768,0.011071999867757162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,128,0.004206222378545337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,64,0.0038684445122877755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,256,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,32,0.0038613333470291565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,256,0.008645333349704742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,128,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,64,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,32,0.008290666672918532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,65536,0.32823554674784344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,65536,0.14272267288631862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,3072,0.025552888711293537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,51200,0.2652631070878771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,65536,0.4299777878655328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,16384,0.08335466517342462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,65536,0.38149510489569766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,12288,0.06299822198020087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,16384,0.04564977685610453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,51200,0.29637244012620717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,512,0.009345778160625035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,10240,0.05310044354862637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,12288,0.03943555553754171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,128,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,12288,0.07260444429185656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,8192,0.04768800073199802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,7168,0.0431600014368693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,10240,0.02609955602222019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,8192,0.021189333664046392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,6144,0.032039112514919706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,7168,0.019613333874278598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,7168,0.053928888506359525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,6144,0.01792977750301361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,6144,0.03994044330384996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,4096,0.020948444803555805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,5120,0.016199111938476562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,5120,0.03435733252101474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,51200,0.11258488231235081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,3584,0.019559999306996662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,4096,0.014144000079896716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,4096,0.028922667105992634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,3584,0.013016000390052795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,16384,0.09543110926946004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,2560,0.014513777361975776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,3584,0.02608000073168013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,3072,0.01203644441233741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,2048,0.012207110722859701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,3072,0.02350844442844391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,2560,0.011035555766688453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,10240,0.061372445689307324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,2560,0.02014133334159851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,1536,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,1024,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,2048,0.010037333601050908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,8192,0.050740443997912936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,768,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,2048,0.017284444636768766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,1536,0.009080000221729279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,512,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,1024,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,5120,0.028088887532552082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,1536,0.014477334088749357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,256,0.0041982221106688184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,768,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,128,0.003549333247873518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,512,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,256,0.007256000406212277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,512,0.00904977818330129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,64,0.0033333332588275275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,32,0.0034942221310403612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,128,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,128,0.007694222033023834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,3072,0.0169831116994222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,65536,0.2460062238905165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,51200,0.18123911486731636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,65536,0.13453866375817194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,65536,0.3377244472503662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,16384,0.07459288835525513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,51200,0.10637688636779785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,1024,0.011388444238238864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,51200,0.27731288803948295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,768,0.010185778141021729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,16384,0.03604355454444885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,10240,0.04576266805330912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,256,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,12288,0.029383109675513372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,8192,0.03598399957021078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,12288,0.06600977977116902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,10240,0.026029333472251892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,64,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,7168,0.030990223089853924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,32,0.00795022232664956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,10240,0.05500177873505486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,6144,0.02626577847533756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,8192,0.04571911030345493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,7168,0.019543111324310303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,5120,0.02273688879277971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,6144,0.017668444249365065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,7168,0.041053334871927895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,4096,0.018664888209766813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,6144,0.03587466809484694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,5120,0.01603200038274129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,3584,0.01667644414636824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,5120,0.03100444542037116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,4096,0.013854222165213691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,3072,0.01440177857875824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,4096,0.02626488937271966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,3584,0.012967999610635968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,2560,0.01276622215906779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,3584,0.023639111055268183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,3072,0.01201333353916804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,12288,0.04929688904020521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,2048,0.010858666565683154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,3072,0.02088266611099243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,16384,0.08445422516928779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,1536,0.00886311133702596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,2560,0.011005333728260465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,1024,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,2560,0.01827822294500139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,1536,0.008971555365456475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,768,0.00564533347884814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,2048,0.015816888875431485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,1536,0.013171555267439948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,1024,0.008386666576067606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,512,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,1024,0.010628444453080496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,768,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,8192,0.021357332666714985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,256,0.0038657776183552216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,128,0.0032115555885765287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,768,0.009655999640623728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,512,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,256,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,64,0.003170666595300039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,128,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,256,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,128,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,64,0.007694222033023834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,65536,0.20536888970269096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,32,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,51200,0.16056177351209852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,65536,0.1316888862186008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,65536,0.2968008783128527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,2048,0.00979466653532452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,16384,0.06491288873884413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,12288,0.04757510953479343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,51200,0.23031110233730742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,51200,0.10569600264231364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,16384,0.034640000926123724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,10240,0.0404151115152571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,16384,0.07463733355204265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,512,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,12288,0.05737421909968058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,10240,0.026664000418451097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,7168,0.028532445430755615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,10240,0.04819733235571119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,32,0.0032302221904198327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,8192,0.02083822256988949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,6144,0.02548444436656104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,8192,0.04015555646684434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,7168,0.019070222973823547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,5120,0.021817778547604878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,7168,0.03666222095489502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,4096,0.01761155492729611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,6144,0.03216266632080078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,5120,0.027631110615200464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,3584,0.01604622271325853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,4096,0.013462222284740873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,4096,0.02307022280163235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,3584,0.012638222012254926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,3584,0.020920889245139226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,3072,0.011523555550310345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,2560,0.012423111332787408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,3072,0.018862222631772358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,12288,0.029345777299669053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,2048,0.009962666365835402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,8192,0.0325582226117452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,1536,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,2560,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,2048,0.009654222263230218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,2560,0.016297777493794758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,1024,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,2048,0.014145778285132514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,1536,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,768,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,1536,0.01205244412024816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,1024,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,512,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,6144,0.017196445001496207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,1024,0.009739555418491364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,768,0.007900444169839224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,256,0.0038737778862317405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,512,0.0074471111098925276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,128,0.0033822223130199644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,768,0.008723555339707268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,64,0.003144888828198115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,256,0.006986666884687211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,512,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,32,0.0032115555885765287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,128,0.007141333487298753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,256,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,3072,0.014135110709402295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,128,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,65536,0.2091697719362047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,64,0.007145778172545963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,51200,0.17621244324578178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,16384,0.05318311187956068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,65536,0.12629600365956625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,51200,0.10000622272491455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,65536,0.2496702141231961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,12288,0.03951911131540934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,16384,0.03380089004834493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,51200,0.19749689102172852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,10240,0.030766222212049697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,12288,0.027853333287768896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,8192,0.03077955709563361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,10240,0.024714666936132643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,12288,0.04911466770701938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,10240,0.04190133346451653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,8192,0.019872888922691345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,5120,0.015525332755512662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,8192,0.03490399983194139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,6144,0.020561777883105807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,7168,0.031722668144438006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,7168,0.018186666899257235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,5120,0.01720977822939555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,6144,0.01643200053109063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,4096,0.013744889034165276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,6144,0.028034667174021404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,5120,0.014720888601409065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,3584,0.013348444468445249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,5120,0.024135111106766596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,4096,0.013070222404268054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,3072,0.011051555474599203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,3584,0.012077333198653327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,3584,0.018258665998776753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,16384,0.06410666969087389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,2560,0.00999022192425198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,3072,0.01129955550034841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,2048,0.008543999658690559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,2560,0.01033333357837465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,7168,0.023044443792766992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,2560,0.014515555567211576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,1536,0.007091555330488417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,2048,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,1024,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,2048,0.012713777522246042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,1536,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,768,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,1024,0.008011555506123437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,1024,0.008964444200197855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,512,0.004171555654870139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,32,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,768,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,256,0.0034888887570963967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,768,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,512,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,4096,0.020093333390023973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,128,0.0028435554769304064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,512,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,64,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,256,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,32,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,128,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,256,0.006755555669466655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,3072,0.016413332687483895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,65536,0.18516000111897787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,64,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,32,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,51200,0.149235553211636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,65536,0.08117510875066121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,16384,0.04712622364362081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,12288,0.03511022196875678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,51200,0.0646488865216573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,65536,0.20392444398668078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,1536,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,16384,0.023064000738991633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,10240,0.0277839998404185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,51200,0.1588080061806573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,8192,0.022335999541812476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,12288,0.019663110375404358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,16384,0.0530426667796241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,7168,0.020222221811612446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,12288,0.04126044445567661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,10240,0.01759199963675605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,6144,0.018017777138286166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,8192,0.014359111587206522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,10240,0.035451554589801364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,7168,0.013435555828942193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,5120,0.015455999308162265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,8192,0.029754665162828233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,4096,0.012895111408498553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,7168,0.027016000615225896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,3584,0.011787555283970304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,5120,0.011184000306659274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,6144,0.023777777949968975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,4096,0.010038221875826517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,5120,0.020632889535692003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,4096,0.016964443855815463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,3584,0.00978577799267239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,2560,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,3584,0.01571644511487749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,3072,0.009394666386975182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,128,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,2048,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,1536,0.006531555619504716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,3072,0.014297778407732645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,2560,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,2048,0.008301333420806462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,1024,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,1536,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,2048,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,768,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,1024,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,1024,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,512,0.0038906666967603895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,768,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,768,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,256,0.0032097777972618737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,512,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,6144,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,512,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,256,0.006525333142942852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,3072,0.010452444354693094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,256,0.00630666646692488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,64,0.0025235555238193935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,32,0.0027991111079851785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,128,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,128,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,65536,0.10528977711995442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,64,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,32,0.0063164445261160536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,2560,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,51200,0.09171555439631145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,16384,0.0320995549360911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,65536,0.07962489128112793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,1536,0.00960000024901496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,65536,0.15583733717600504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,51200,0.06425333023071289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,16384,0.022465777066018846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,51200,0.12343911329905193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,10240,0.025087111526065405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,12288,0.0188071115149392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,8192,0.02239822182390425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,12288,0.032997333341174655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,7168,0.018182223041852314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,10240,0.017456889152526855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,10240,0.02854222059249878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,8192,0.013814222481515674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,128,0.002871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,6144,0.017299556069903903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,8192,0.024159999357329473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,7168,0.012734221915404001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,7168,0.021725333399242822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,4096,0.010799999866220685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,5120,0.010732444624106089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,6144,0.01888444523016612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,3584,0.008828444613350762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,5120,0.0166657782263226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,4096,0.009822222093741098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,4096,0.014106666048367819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,3584,0.009550222092204625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,3584,0.012842666771676807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,2560,0.007034666836261749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,3072,0.009147555463843876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,12288,0.03022222386466132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,2048,0.00620177760720253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,3072,0.011757333245542316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,2560,0.008622222476535374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,1536,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,16384,0.04223733478122287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,2560,0.010775110787815519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,2048,0.008174222376611497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,1024,0.0041759999261962045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,2048,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,1536,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,768,0.0038515557017591265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,1024,0.0070168889231152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,1536,0.008642666869693333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,512,0.0032088889016045462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,5120,0.013748444616794586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,1024,0.007105777661005656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,768,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,256,0.0025920000755124623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,6144,0.01184800018866857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,128,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,512,0.00657066661450598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,768,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,64,0.0021671110557185277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,256,0.006310222049554189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,32,0.0024542222834295696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,512,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,128,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,256,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,65536,0.0950195524427626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,128,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,3072,0.008034666379292807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,64,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,32,0.00572533326016532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,16384,0.029567112525304157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,12288,0.027514666318893433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,65536,0.1543288893169827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,16384,0.021942221456103857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,51200,0.1212115552690294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,10240,0.023356444305843774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,16384,0.041344000233544245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,12288,0.018329777651362948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,8192,0.019887111253208585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,10240,0.016774222254753113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,12288,0.03243466549449497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,7168,0.0166311115026474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,10240,0.028068443139394123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,8192,0.013524444566832649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,8192,0.023323555787404377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,6144,0.013721778160995908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,7168,0.01292888902955585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,5120,0.011633777783976661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,7168,0.021291555629836187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,6144,0.018844443890783522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,4096,0.009389333426952362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,5120,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,5120,0.016521778371598985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,51200,0.08116088973151313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,3584,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,4096,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,65536,0.07907111114925809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,3072,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,4096,0.014105778601434497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,51200,0.06363111072116427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,2560,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,3584,0.012791111237472959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,3072,0.008729777402347988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,2048,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,2560,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,3072,0.011756444142924415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,1536,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,2560,0.010695999695195092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,1024,0.00444088876247406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,2048,0.008056888977686564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,1536,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,2048,0.009574222067991892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,768,0.003866666720973121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,1536,0.008349333372380998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,512,0.003206222214632564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,768,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,6144,0.011749332977665795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,256,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,768,0.006572444405820634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,512,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,128,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,512,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,256,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,64,0.0023182222826613323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,128,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,256,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,32,0.0022106667359670005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,128,0.005412444472312927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,65536,0.061418665779961475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,3584,0.009378666679064432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,64,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,32,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,65536,0.07902577850553724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,16384,0.02125599980354309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,65536,0.11114488707648383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,51200,0.06359733475579156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,12288,0.017684444785118103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,51200,0.08760622474882339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,16384,0.021191999316215515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,1024,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,10240,0.015674667225943673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,16384,0.031214220656288996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,1024,0.007063111497296228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,12288,0.0181368887424469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,10240,0.016474665866957772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,7168,0.011753777662913004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,12288,0.024726221958796184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,8192,0.013442666994200813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,10240,0.021735999319288466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,8192,0.018064889642927382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,7168,0.012617777619096967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,5120,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,6144,0.011709333293967776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,4096,0.007376888559924231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,6144,0.01494488947921329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,5120,0.013425778183672162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,3584,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,51200,0.04526933365397983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,4096,0.01166133334239324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,3072,0.006169777777459886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,3584,0.009311111436949836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,3584,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,2560,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,3072,0.008958222137557136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,3072,0.00980711148844825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,2048,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,8192,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,2560,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,1536,0.004173333446184794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,2560,0.009154666629102495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,6144,0.00998222248421775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,2048,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,1024,0.00349866681628757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,1536,0.00738666703303655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,1536,0.00742222203148736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,7168,0.016608888904253643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,1024,0.0070257774657673305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,512,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,5120,0.010758221977286868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,1024,0.0063075555695427795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,768,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,256,0.002499555548032125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,512,0.006308444258239534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,512,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,128,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,256,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,256,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,64,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,128,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,32,0.0021226666867733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,128,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,65536,0.04173511266708374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,64,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,2048,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,32,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,51200,0.03285155693689982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,65536,0.07856266366110908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,65536,0.10926133394241333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,16384,0.014818666709793938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,51200,0.06269333097669813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,51200,0.08622044324874878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,12288,0.012255111502276527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,16384,0.021017778250906203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,10240,0.011092444260915121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,16384,0.030324445830451116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,4096,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,12288,0.01779733267095354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,8192,0.009847111172146266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,12288,0.02400355537732442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,10240,0.016224000189039443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,7168,0.01068800025516086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,8192,0.013436444103717804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,10240,0.020787555310461257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,6144,0.009646221995353699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,8192,0.017839110559887357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,5120,0.008310221963458592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,7168,0.012437333663304647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,4096,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,7168,0.01626399987273746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,6144,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,3584,0.006325333482689328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,5120,0.0107351111041175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,6144,0.014714666538768344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,3072,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,5120,0.013180444637934366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,4096,0.009678222239017487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,2560,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,3584,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,4096,0.011373333632946014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,3072,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,3584,0.010664888554149203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,768,0.0031582222630580268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,2560,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,3072,0.009723555710580613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,1536,0.0038364442686239877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,1024,0.0032026666320032547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,2560,0.009115555220180089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,2048,0.007956444389290279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,768,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,2048,0.008348444269763099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,1024,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,768,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,1536,0.007348444726732042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,512,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,256,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,768,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,1024,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,128,0.0021102221475707158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,512,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,768,0.005982222242487802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,64,0.0019466665883858998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,256,0.006217777729034424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,512,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,128,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,256,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,65536,0.0326186650329166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,128,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,64,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,51200,0.026086222794320848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,32,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,65536,0.0783626635869344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,16384,0.011728888584507836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,65536,0.10870755381054348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,51200,0.06236622068617079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,2048,0.00453333349691497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,16384,0.02088444431622823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,51200,0.08579466740290324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,10240,0.008829333715968663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,16384,0.02961955467859904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,12288,0.017747556169827778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,8192,0.008175111479229396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,10240,0.016224000189039443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,1536,0.007692444655630324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,7168,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,12288,0.023623110519515142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,10240,0.020623111062579684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,6144,0.0070257774657673305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,8192,0.013189333180586496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,7168,0.012441777520709567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,8192,0.01757333344883389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,6144,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,7168,0.016227554943826463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,4096,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,6144,0.014489778214030795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,5120,0.010611555642551847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,5120,0.013089777694808112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,3584,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,4096,0.009686222506894005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,3072,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,4096,0.011423110961914062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,3584,0.009167999857001835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,2560,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,3584,0.01053155544731352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,3072,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,12288,0.009754666851626502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,2048,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,3072,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,2560,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,1536,0.003858666867017746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,2048,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,2560,0.008972444468074376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,1024,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,1536,0.007664889097213745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,2048,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,768,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,1536,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,1024,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,512,0.0024844445288181305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,5120,0.0074177781740824384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,768,0.006675555474228329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,1024,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,256,0.002199999988079071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,768,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,512,0.006247111078765657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,256,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,512,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,32,0.002130666747689247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,256,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,128,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,65536,0.029576000240114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,128,0.005203555441564984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,51200,0.02299822204642826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,16384,0.010441777606805166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,64,0.004984000076850255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,12288,0.009364444348547194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,32,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,10240,0.008655110994974772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,65536,0.10839555660883586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,8192,0.007575111256705389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,16384,0.029490666257010564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,12288,0.023398222194777593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,7168,0.007072000039948358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,6144,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,10240,0.02040799955526988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,5120,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,7168,0.01621866722901662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,6144,0.014510222607188754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,4096,0.006911110960774952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,3584,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,5120,0.012868444124857584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,32,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,3072,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,2560,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,2048,0.004209777961174647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,4096,0.011398221883508893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,3584,0.010400888820489248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,1536,0.0038968887594011095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,1024,0.0031999999450312722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,3072,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,2560,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,768,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,2048,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,512,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,256,0.0021964444054497615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,1536,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,1024,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,128,0.002128888956374592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,64,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,768,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,32,0.0019066666977273093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,512,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,256,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,65536,0.024663110574086506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,128,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,64,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,16384,0.016024000114864774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,32,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,12288,0.01036088913679123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,128,0.002116444417172008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,65536,0.10766310824288262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,10240,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,51200,0.08542755577299331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,8192,0.008643555144468943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,16384,0.029350221157073975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,7168,0.013376889129479727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,6144,0.01146666705608368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,12288,0.023354666100607976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,10240,0.020593777298927307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,8192,0.01741066740618812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,7168,0.016181333197487723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,4096,0.008548444343937768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,6144,0.01443288889196184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,3584,0.007803555164072249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,51200,0.08579911126030816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,5120,0.012921777864297232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,4096,0.011398221883508893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,2560,0.006155555446942647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,3584,0.010399999717871347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,8192,0.01750577820671929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,2048,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,1536,0.004471111214823193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,3072,0.009720889230569204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,1024,0.003524444583389494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,2560,0.008992888861232335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,768,0.0031964443624019623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,2048,0.007985778152942657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,1536,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,512,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,256,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,768,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,5120,0.009559999737474654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,128,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,512,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,256,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,32,0.0019262221952279408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,128,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,64,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,3072,0.006607111129495833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,32,0.004947555561860402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,51200,0.018052443861961365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,1024,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,64,0.0018764444523387486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,16384,1.2310079998440213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,16384,0.4156720108456082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,16384,1.4206790924072266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,12288,0.9157955381605359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,51200,1.4348684946695964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,12288,0.32116354836357963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,12288,1.0539831585354276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,51200,3.8530061509874134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,10240,0.910056855943468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,16384,1.2107582092285156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,10240,0.8291715515984429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,8192,0.7365848753187391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,12288,0.8644711176554362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,10240,0.2746968799167209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,51200,4.6709980434841585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,8192,0.6875351270039877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,8192,0.22258933385213217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,7168,0.5524640083312988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,10240,0.7220995691087512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,7168,0.19680710633595785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,7168,0.5885519981384277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,6144,0.5413422054714626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,6144,0.17234222094217935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,8192,0.5760604540506998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,5120,0.3631617758009169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,6144,0.5680658022562662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,7168,0.5049955579969618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,4096,0.2981404463450114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,5120,0.14797955089145237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,5120,0.6122053464253744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,4096,0.3337528970506456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,3584,0.2780017852783203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,4096,0.13049156135982937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,5120,0.36805955568949383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,3584,0.3237733311123318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,4096,0.3011173407236735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,3584,0.11578222115834554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,2560,0.21333866649203828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,3072,0.0976311100853814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,3072,0.2769164509243435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,3584,0.2646142111884223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,2048,0.15672177738613552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,2560,0.08549955818388198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,3072,0.23574667506747773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,2560,0.21490666601392958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,6144,0.4342248969607883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,1536,0.1375653346379598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,2560,0.19970756106906465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,2048,0.172179553243849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,1024,0.0778684417406718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,1536,0.13619466622670492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,3072,0.23267110188802084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,2048,0.16320710712009007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,768,0.07171289126078288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,1024,0.04646933409902784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,1536,0.13031821780734593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,1024,0.10363822513156468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,512,0.04632622334692213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,768,0.0837999979654948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,768,0.04144800040456984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,1024,0.10003644227981567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,256,0.034705777963002525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,512,0.037754668129815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,768,0.08425244357850815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,51200,3.9340231153700085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,128,0.040101332796944514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,2048,0.0733671122127109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,256,0.034648888640933566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,512,0.06895999775992499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,256,0.06573422087563409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,64,0.026107556290096704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,1536,0.05794577466117012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,128,0.03405333227581448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,256,0.05432977941301134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,32,0.02641244398223029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,64,0.060553775893317334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,128,0.04207911094029745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,32,0.060885330041249595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,512,0.07196533017688327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,128,0.06072533130645752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,65536,1.446567111545139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,51200,3.1794471740722656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,65536,4.083065880669488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,16384,1.202381345960829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,51200,3.710593753390842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,51200,1.098264906141493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,65536,4.869434780544705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,16384,1.1085555818345811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,16384,0.36296534538269043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,12288,0.7247920036315918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,12288,0.25337065590752494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,10240,0.7034088770548502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,12288,0.9637716081407336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,10240,0.6665075620015463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,8192,0.4819555812411838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,10240,0.24453510178460014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,16384,0.8858462439643012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,12288,0.683319091796875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,8192,0.17655732896592882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,8192,0.5158320003085667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,7168,0.4508675469292535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,7168,0.15653777122497559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,6144,0.4083048767513699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,8192,0.4489626884460449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,6144,0.385225772857666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,6144,0.13606754938761392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,7168,0.4015671147240533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,10240,0.5567626423305935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,5120,0.27960888544718426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,51200,3.1046346028645835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,4096,0.22486488024393717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,5120,0.11754666434393989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,7168,0.46777598063151044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,4096,0.27330756187438965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,5120,0.29161421457926434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,3584,0.21474399831559923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,65536,3.9413350423177085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,3584,0.08774844143125747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,4096,0.23571822378370497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,3584,0.23162134488423666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,3072,0.1839404503504435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,3584,0.21210666497548422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,3072,0.0787315567334493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,2560,0.15499022271898058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,3072,0.19308355119493273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,5120,0.3203084468841553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,2048,0.12699466281467015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,3072,0.18359643883175322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,2560,0.16564355956183538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,4096,0.09789599974950154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,1536,0.08909155262841119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,2048,0.13593244552612305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,2560,0.15955732928382024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,1024,0.07396622498830159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,1536,0.045688887437184654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,2048,0.13079022036658391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,1536,0.10833688577016194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,768,0.05252888798713684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,1024,0.08252533276875813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,1024,0.0373253325621287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,1536,0.10633866654502021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,512,0.04107644491725498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,768,0.03424177898301019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,768,0.06697689162360297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,1024,0.08083911074532403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,6144,0.3472417725457086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,256,0.02849688794877794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,512,0.031111111243565876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,2560,0.06891466511620416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,512,0.05761333306630453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,128,0.022287110487620037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,256,0.02858933475282457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,256,0.052582220898734204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,512,0.056186669402652316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,2048,0.057244446542527944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,64,0.0199297782447603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,32,0.02055911057525211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,256,0.04545866780810886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,128,0.03506133291456435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,64,0.04866844415664673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,32,0.04885066548983256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,768,0.06874933507707384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,65536,0.47804535759819883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,65536,1.2614542643229167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,128,0.04888977938228183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,51200,1.0424657397800021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,16384,0.3037750985887316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,65536,1.6131715774536133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,51200,0.36610134442647296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,16384,0.33554932806226945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,51200,1.183416896396213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,16384,0.11324889130062527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,12288,0.23780979050530326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,128,0.02812355425622728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,16384,0.2971137894524468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,12288,0.26206310590108234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,12288,0.08882311317655776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,10240,0.21441600057813856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,10240,0.07618133227030437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,12288,0.22447734408908418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,65536,1.2277332941691081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,8192,0.15378933482699925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,7168,0.13459822866651747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,8192,0.07043644454744127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,51200,0.9485093222724067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,8192,0.15339377191331652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,7168,0.06514488988452487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,6144,0.1193279955122206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,10240,0.22477245330810547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,7168,0.14709688557518855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,5120,0.10642044411765204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,7168,0.13719111018710667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,6144,0.04787110951211718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,6144,0.12770844830407038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,4096,0.08230666981803046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,5120,0.1079839997821384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,5120,0.04181066817707486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,6144,0.12081066767374675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,3584,0.07544177770614624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,4096,0.035007112556033664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,4096,0.08735377920998467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,8192,0.17075555854373506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,3072,0.05883200301064385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,3584,0.030348443322711523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,4096,0.08686577611499363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,2560,0.05043111244837443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,3072,0.02732533382044898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,3584,0.07671911186642118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,3072,0.06909066438674927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,2048,0.04756977823045519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,2560,0.025248888466093276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,3072,0.06833599673377143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,2560,0.05940888987647163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,1536,0.03280977739228143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,2048,0.04921333326233757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,2048,0.021886222892337378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,2560,0.059939556651645236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,1024,0.024136889312002394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,1536,0.018805333309703402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,2048,0.051649779081344604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,1536,0.03927910990185208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,768,0.018550222118695576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,5120,0.10213066471947564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,1024,0.03051466743151347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,1536,0.0433937774764167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,1024,0.015633778439627755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,512,0.014140443669425117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,768,0.014031999640994601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,768,0.02516533268822564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,256,0.0107004443804423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,512,0.012906666431162091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,512,0.02107644412252638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,768,0.030089779032601252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,128,0.009329777624871995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,512,0.025880000657505457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,256,0.011972444752852121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,256,0.01856355534659492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,64,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,128,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,32,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,128,0.01812266641192966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,256,0.021111999948819477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,64,0.01789955629242791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,128,0.01686844395266639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,32,0.01792977750301361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,10240,0.19360444280836317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,3584,0.07772799995210436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,1024,0.034295111894607544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,65536,0.3675093385908339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,65536,0.9353564580281576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,51200,0.7380675739712186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,16384,0.27035734388563365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,65536,1.1638995276557076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,51200,0.288062228096856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,51200,0.8337919976976184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,16384,0.282134215037028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,12288,0.20142666498819986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,10240,0.1670302285088433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,12288,0.07010755274030897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,12288,0.18896978431277803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,16384,0.2641857730017768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,8192,0.13707644409603542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,10240,0.06052088737487793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,12288,0.18879821565416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,10240,0.1747839980655246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,51200,0.7783066431681315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,10240,0.15906666384802923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,8192,0.142666671011183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,65536,0.995031992594401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,8192,0.12877955701616076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,7168,0.047897779279285006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,16384,0.09225244654549493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,7168,0.11414044433169895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,7168,0.11874843968285455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,6144,0.038741333617104426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,6144,0.10329955816268921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,5120,0.08430933290057713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,5120,0.033486220571729876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,6144,0.10326311323377822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,5120,0.08442133002811009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,4096,0.07071822219424777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,7168,0.11933244599236383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,5120,0.08868266476525201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,4096,0.02858577834235297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,3584,0.06150666872660319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,4096,0.06894310977723864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,3072,0.05819911426968045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,3584,0.02581688927279578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,4096,0.07413422399097018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,3584,0.0612693362765842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,6144,0.09609778059853448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,2560,0.04340089029735989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,3072,0.023696889479955036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,3072,0.054117331902186074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,3584,0.06609244479073419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,2048,0.03764177693261041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,2560,0.02139644490347968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,2560,0.04644533329539829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,1536,0.02551377813021342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,2048,0.01862044466866387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,2560,0.05217511124081082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,2048,0.045125332143571645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,1536,0.016158221496476066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,1024,0.01889600025282966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,8192,0.050006220738093056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,1536,0.03112799922625224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,768,0.01511111193233066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,1536,0.03762222329775492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,1024,0.013763555222087435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,1024,0.02442577812406752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,512,0.011683555940786997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,768,0.012680889003806643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,1024,0.030125333203209773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,768,0.020422221885787118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,256,0.011340444286664328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,512,0.016984888248973422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,512,0.011388444238238864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,768,0.026342221432262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,128,0.009462222456932068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,512,0.022316444251272414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,3072,0.058879110548231334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,256,0.015159999330838522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,64,0.00888888869020674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,32,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,2048,0.03849511014090644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,256,0.018186666899257235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,128,0.01483022173245748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,64,0.014767999450365702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,128,0.015156444576051501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,32,0.014783999986118741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,65536,0.8505911297268338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,65536,0.3234248956044515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,51200,0.6624435318840874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,256,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,65536,0.9166533152262369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,16384,0.23000621795654297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,51200,0.7047137684292264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,128,0.01091288857989841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,12288,0.1821742190255059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,51200,0.24889689021640352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,16384,0.08165599902470906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,16384,0.2196568912929959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,12288,0.1618071132236057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,12288,0.061863998572031655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,16384,0.21301510598924425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,10240,0.05299377772543165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,12288,0.16467644108666316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,65536,0.891351064046224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,10240,0.1353751156065199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,8192,0.04799377918243408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,8192,0.11355733871459961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,7168,0.11740711000230576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,51200,0.6541324191623265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,8192,0.11234222518073188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,6144,0.08157777786254883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,7168,0.0373617774910397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,7168,0.10559288660685222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,5120,0.08095555835300021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,6144,0.03330755564901564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,6144,0.08430666393703884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,7168,0.09996800290213691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,4096,0.055431998438305326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,5120,0.02920088834232754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,5120,0.07288089063432482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,6144,0.09561155239741008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,10240,0.13618222872416177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,3584,0.05424088901943631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,4096,0.025095999240875244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,5120,0.07543555233213636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,4096,0.05912444326612684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,3072,0.04106222258673774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,3584,0.0226248883538776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,3584,0.052707556221220225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,4096,0.06392355759938557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,2560,0.03909777932696872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,3072,0.02088177700837453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,3072,0.046408888366487294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,3584,0.058338668611314565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,2048,0.03177155719863044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,3072,0.05129955543412102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,2560,0.040242665343814425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,2560,0.01900622248649597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,1536,0.024803555674023096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,10240,0.15632532702551946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,2048,0.016751999656359356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,2048,0.03404266635576884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,1024,0.01646222174167633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,8192,0.108824888865153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,1536,0.01387111097574234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,1536,0.02760444415940179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,768,0.013269333375824822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,1024,0.021358221769332886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,1536,0.03335288829273648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,768,0.011326221956147088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,512,0.01032266683048672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,768,0.023733332753181458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,512,0.015228443675571017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,512,0.010346666806273991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,256,0.0074444446298811175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,2560,0.04561511013242933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,512,0.01978933314482371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,2048,0.0387697782781389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,256,0.009999111294746399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,128,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,1024,0.012100444071822695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,256,0.01719199948840671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,64,0.007375111182530721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,128,0.009916444619496664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,32,0.007551111280918121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,768,0.01793155570824941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,128,0.01310400002532535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,1024,0.02676800058947669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,128,0.013799111048380533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,64,0.013103110922707452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,32,0.013145777914259167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,65536,0.2538462214999729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,65536,0.5922026634216309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,51200,0.5400746663411459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,16384,0.17087111208173963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,65536,0.7374258041381836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,256,0.013412444127930535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,51200,0.19865689012739393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,51200,0.5815848774380153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,12288,0.13754578431447348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,16384,0.19695999887254503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,16384,0.06474133332570393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,10240,0.11327466699812148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,12288,0.05073777834574381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,12288,0.13591021961636013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,16384,0.18221688270568848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,51200,0.5429484579298232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,65536,0.6945199966430664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,8192,0.09371111128065322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,12288,0.13513955805036756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,7168,0.08151377571953668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,8192,0.03979466689957513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,8192,0.09209333525763617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,10240,0.11436088879903157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,6144,0.06988800234264798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,7168,0.08020444711049397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,7168,0.03545066714286804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,8192,0.09666577974955241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,5120,0.05903555287255181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,6144,0.028499556912316218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,6144,0.0702471137046814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,7168,0.08655911021762425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,4096,0.0468444459968143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,5120,0.06023110946019491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,5120,0.025067556235525344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,6144,0.07511911127302381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,3584,0.037104000647862755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,4096,0.04896177848180135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,4096,0.02144533395767212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,5120,0.06686222553253174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,3072,0.03601866629388597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,3584,0.01959822244114346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,3584,0.0431600014368693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,10240,0.11151199870639378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,10240,0.04337955514589945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,2560,0.02938133478164673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,3072,0.038712001509136625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,2048,0.024461333950360615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,2560,0.016551999582184684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,3072,0.04451644420623779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,2560,0.033928000264697604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,1536,0.019188443819681805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,2560,0.03952177696757846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,2048,0.028559111886554297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,2048,0.014839111102951897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,2048,0.034083555142084755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,1536,0.012783110969596438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,1024,0.014208889669842191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,1536,0.02312266661061181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,1024,0.011188444164064197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,1536,0.028868443436092798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,768,0.011584888729784222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,1024,0.01826755536927117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,4096,0.05409155620468987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,1024,0.02365066607793172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,768,0.010088000032636855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,512,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,768,0.015344889627562629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,3072,0.01812000075976054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,256,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,512,0.009372444616423713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,3584,0.05019111103481717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,768,0.020987555384635925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,512,0.013117333253224691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,256,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,256,0.011747555600272285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,512,0.018173333671357896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,64,0.006217777729034424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,32,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,128,0.009141333401203156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,128,0.011390222443474663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,256,0.015840000576443143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,64,0.011382222175598145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,128,0.012739555703269111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,32,0.011353777514563667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,65536,0.6024986902872721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,65536,0.2362284395429823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,51200,0.46744977103339297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,65536,0.6538177596198188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,16384,0.15551555156707764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,128,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,51200,0.5139297909206814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,51200,0.18586311075422499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,12288,0.11709245045979817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,16384,0.0602746672100491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,16384,0.1628684467739529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,12288,0.04673422376314799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,10240,0.10895288652843899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,12288,0.11931999524434407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,65536,0.6842631234063042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,16384,0.1747137837939792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,10240,0.04031733340687222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,12288,0.13427644305759007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,10240,0.10198844141430324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,7168,0.07631733020146687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,8192,0.03320711188846164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,51200,0.5368142127990723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,8192,0.08915288580788507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,10240,0.11440800295935737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,6144,0.06769955820507474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,7168,0.02986577815479702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,7168,0.07283733288447063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,8192,0.09335111247168647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,6144,0.027958220905727808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,6144,0.06400266620847914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,7168,0.08384355571534897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,4096,0.04426844583617317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,5120,0.024860444996092055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,5120,0.054213331805335156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,6144,0.07450311051474677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,3584,0.037124445041020714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,4096,0.020738666256268818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,4096,0.04469688733418783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,5120,0.06324622366163465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,3072,0.03334044416745504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,3584,0.04035733474625482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,8192,0.0884515576892429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,4096,0.05399466554323832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,2560,0.029269334342744615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,3072,0.017601778109868366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,3584,0.04846577843030294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,3072,0.03574488891495599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,2048,0.023735110958417256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,2560,0.015856888559129503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,2560,0.0316826668050554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,3072,0.043459554513295494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,1536,0.019134221805466544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,2560,0.03853866789076064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,5120,0.0558240016301473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,1024,0.01460088955031501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,1536,0.012310222619109683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,2048,0.03260711166593764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,1536,0.028064890040291682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,1024,0.016534222496880423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,768,0.01202044470442666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,1024,0.010681777364677854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,768,0.00999466660949919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,3584,0.019147555033365887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,512,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,768,0.014312000738249885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,768,0.01958933307064904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,512,0.009321778184837764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,256,0.006664000037643645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,512,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,512,0.017497777938842773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,2048,0.013888888888888888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,256,0.01108088923825158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,128,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,2048,0.026486222942670185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,256,0.00886577781703737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,64,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,1536,0.021741333934995864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,256,0.014440000057220459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,32,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,128,0.008989333278603023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,128,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,64,0.010692444112565784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,32,0.010672888822025724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,128,0.012183999849690331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,1024,0.022934221559100684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,65536,0.5262204276190864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,65536,0.22587821218702528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,51200,0.41451112429300946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,65536,0.563660462697347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,16384,0.13480622238583034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,51200,0.4556364483303494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,51200,0.21005333794487846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,16384,0.1500097778108385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,12288,0.0866968896653917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,65536,0.5380044513278538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,12288,0.040828443235821195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,12288,0.11644444200727676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,51200,0.4248017734951443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,10240,0.09581866529252793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,8192,0.07487199703852336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,10240,0.03825777769088745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,12288,0.10951822333865696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,10240,0.08917778068118626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,7168,0.056792888376447886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,8192,0.028652442826165095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,8192,0.07678933276070489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,10240,0.09365422195858425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,6144,0.056384887960222035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,7168,0.027595554788907368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,7168,0.06442311074998644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,8192,0.07802310917112562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,5120,0.04184800055291918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,6144,0.02389777700106303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,16384,0.05489333470662435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,6144,0.056141330136193164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,7168,0.0682800014813741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,4096,0.037760890192455716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,5120,0.021589333812395733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,5120,0.04854933420817057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,6144,0.061208003097110324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,3584,0.03381244341532389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,16384,0.14046400123172337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,4096,0.03961866762903001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,3072,0.029088000456492107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,5120,0.054217778974109225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,3584,0.016907556189431083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,2560,0.025161777933438618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,4096,0.044870220952563815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,3072,0.015442666080262927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,3072,0.031856000423431396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,2048,0.0203155560625924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,3584,0.040896889236238264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,2560,0.028318223026063707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,2560,0.014039110806253223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,3072,0.03690222236845229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,1536,0.014183999763594734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,2560,0.03272355596224467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,2048,0.023794665932655334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,1024,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,1536,0.01109777804878023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,1536,0.01941511034965515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,1536,0.024526221884621516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,1024,0.015069334043396844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,1024,0.009580444130632613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,768,0.010053333308961656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,4096,0.01812622282240126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,512,0.008174222376611497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,768,0.009012444151772393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,1024,0.01943199998802609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,768,0.013134222063753339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,3584,0.0358533329433865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,256,0.0064533332155810455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,512,0.008624888956546783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,768,0.017940445078743827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,128,0.0058266665372583605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,256,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,512,0.016895999511082966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,256,0.01481066644191742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,128,0.0083137775460879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,2048,0.012609778179062737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,32,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,2048,0.02854666776127285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,128,0.011416888899273343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,64,0.009832888841629028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,32,0.009740444521109263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,65536,0.37574489911397296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,65536,0.18195645014444986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,512,0.011054221954610614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,51200,0.30568088425530326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,65536,0.48883819580078125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,256,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,16384,0.1188017792171902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,64,0.005375999957323074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,128,0.009672000176376766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,51200,0.3713360097673204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,65536,0.5290115674336752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,12288,0.0847804413901435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,16384,0.11799821588728164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,12288,0.038133333126703896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,16384,0.1395448843638102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,51200,0.41569333606296116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,12288,0.10798222488827175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,10240,0.07572711176342435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,8192,0.060053335295783154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,10240,0.03343377841843499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,10240,0.09183555841445923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,51200,0.1444177760018243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,7168,0.053483555714289345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,8192,0.061807109249962695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,8192,0.027178666657871667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,16384,0.04776888754632738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,8192,0.07573511203130086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,7168,0.024847111768192712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,6144,0.04394666684998406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,7168,0.055847108364105225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,12288,0.09247289101282756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,6144,0.022452443838119507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,10240,0.07726399766074286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,7168,0.06764533122380574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,6144,0.04810755451520284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,4096,0.029898666673236426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,5120,0.0199937770764033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,6144,0.05992177459928724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,5120,0.041745778587129384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,3584,0.02306844459639655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,4096,0.01756000022093455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,4096,0.03452355662981669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,3072,0.022691556149058874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,5120,0.05184088812934028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,3584,0.0159244438012441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,4096,0.04416177670160929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,3584,0.031145777967241075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,2560,0.019179556104871962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,3072,0.014544889330863953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,3072,0.027678221464157104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,2048,0.015885333220163982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,2560,0.013472889032628802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,2560,0.024653333756658766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,3072,0.03605422377586365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,2048,0.012115555504957834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,2560,0.03242311212751601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,2048,0.020657777786254883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,1024,0.008736000292830998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,5120,0.03749955693880717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,2048,0.027715555495686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,1536,0.010410666465759277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,1536,0.01722755531469981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,768,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,1024,0.009327111144860586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,1024,0.01331377774477005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,512,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,1536,0.02346222268210517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,768,0.008955555657545725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,1024,0.01904622217019399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,768,0.011701333026091257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,256,0.005196444276306364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,512,0.008340444829728868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,768,0.01790577835506863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,128,0.004552888787455029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,256,0.00795377790927887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,256,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,3584,0.040044443474875555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,512,0.015477332803938123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,64,0.004312889029582341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,32,0.004519999855094486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,256,0.013591110706329346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,128,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,1536,0.01236533373594284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,64,0.00906044410334693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,32,0.00903555585278405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,128,0.011355555719799466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,65536,0.3102017773522271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,65536,0.13967377609676784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,51200,0.2306177881028917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,65536,0.3977128929562039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,16384,0.07570666737026639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,51200,0.3145502143436008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,51200,0.11111911137898763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,512,0.010015111002657149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,65536,0.3880017863379584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,16384,0.037308444579442344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,16384,0.09735466374291314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,51200,0.30604712168375653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,10240,0.04933599962128533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,12288,0.029614223374260798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,128,0.009022222624884712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,16384,0.10618844297197129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,8192,0.04568799999025133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,10240,0.02604266670015123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,10240,0.0638435549206204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,12288,0.0839937792883979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,7168,0.03825155562824673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,8192,0.0517537792523702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,10240,0.07028444608052571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,8192,0.021908443835046556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,6144,0.036355555057525635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,7168,0.01999200052685208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,7168,0.04626489016744825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,8192,0.05829333596759372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,5120,0.02973777718014187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,6144,0.01827022267712487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,7168,0.0528035561243693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,12288,0.07069955269495647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,4096,0.02514311174551646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,5120,0.035016000270843506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,5120,0.016536000702116225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,6144,0.046985778543684215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,12288,0.07425066497590807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,3584,0.021430222524536982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,4096,0.014327999618318347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,5120,0.04199644592073229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,3072,0.01926933394538032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,3584,0.013166222307417126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,4096,0.03514400124549866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,3584,0.026322666141721938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,2560,0.016695999436908297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,3072,0.023765333824687537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,3072,0.012094222009181976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,3584,0.03234400020705329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,2048,0.014301333162519666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,2560,0.010960000256697336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,2560,0.020776000287797716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,3072,0.029492444462246362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,2048,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,6144,0.040617777241600886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,2560,0.026320000489552815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,1024,0.008643555144468943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,1536,0.009096889032257928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,2048,0.02329155637158288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,1024,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,4096,0.029515554507573444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,1024,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,768,0.007927999728255803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,768,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,1024,0.01650044487582313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,512,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,768,0.010397333237859938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,768,0.015154666370815702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,512,0.007632888853549957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,512,0.009014222357008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,1536,0.011711111499203576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,512,0.013819555441538492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,2048,0.017635555730925668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,256,0.0086986662613021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,256,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,1536,0.014521777629852295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,256,0.01240799989965227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,128,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,32,0.004138666722509596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,128,0.008275555239783393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,1536,0.019848888119061787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,128,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,64,0.008276444342401292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,32,0.008291555775536431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,65536,0.2916719913482666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,51200,0.22914576530456543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,65536,0.13550844457414415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,256,0.004628444297446145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,65536,0.3749528990851508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,16384,0.08344711197747125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,51200,0.29155556360880536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,51200,0.10733244154188369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,128,0.0041759999261962045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,12288,0.06531111399332683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,65536,0.3840204344855414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,16384,0.03674133287535773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,16384,0.0920577777756585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,10240,0.053972442944844566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,64,0.003920000046491623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,12288,0.029967112673653498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,8192,0.044681777556737266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,12288,0.0711866683430142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,51200,0.3035919931199816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,10240,0.02623288830121358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,10240,0.06018666426340739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,7168,0.03892622060245938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,12288,0.0819937785466512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,8192,0.021551999780866835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,6144,0.034620443979899086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,8192,0.05021599928538004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,10240,0.06983200046751234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,7168,0.04400711258252462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,5120,0.028696888022952612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,8192,0.05810844235950046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,6144,0.03924177752600776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,7168,0.05240355597601997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,4096,0.02370577719476488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,5120,0.016213332613309223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,6144,0.047056890196270414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,5120,0.03372711274358962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,4096,0.014244443840450711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,5120,0.04065600037574768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,3584,0.0214035560687383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,4096,0.028498666154013738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,16384,0.10560533073213364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,4096,0.035303112533357404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,3584,0.01294222225745519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,3584,0.02569599946339925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,3584,0.03177777926127116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,7168,0.019877332780096266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,3072,0.023018666439586218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,3072,0.011968000067604912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,6144,0.018073777357737224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,2560,0.010671111444632212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,2048,0.014239999983045789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,2560,0.02022133270899455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,2560,0.026349332597520616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,2048,0.009679999616410997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,1536,0.011778666741318174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,2048,0.016920889417330425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,2048,0.022240888741281297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,1536,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,1024,0.00903200027015474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,1536,0.014263111684057446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,3072,0.019342222147517733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,768,0.007556444240940942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,1024,0.008295999632941352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,1536,0.019446222318543326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,1024,0.011258666714032492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,512,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,768,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,1024,0.016541333662139047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,768,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,256,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,512,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,3072,0.028944889704386394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,768,0.0151573336786694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,512,0.00869511150651508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,128,0.004173333446184794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,256,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,512,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,64,0.003882666842805015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,32,0.004229333251714706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,128,0.007720000214046902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,128,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,256,0.012103999654452005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,128,0.010076444182131026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,65536,0.23644532097710502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,32,0.00775111135509279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,51200,0.1847235494189792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,65536,0.1302080021964179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,65536,0.32774578200446236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,2560,0.01647200021478865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,51200,0.2555813259548611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,51200,0.10353955295350815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,65536,0.3846293290456136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,16384,0.08295110861460368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,256,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,51200,0.3053013218773736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,16384,0.1047839985953437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,12288,0.06379910972383288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,10240,0.04638933473163181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,64,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,12288,0.028804444604449805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,12288,0.08170755704243977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,10240,0.025807999902301367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,10240,0.05407289001676771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,16384,0.06902044349246554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,8192,0.04494844542609321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,7168,0.032163557079103254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,8192,0.05858133236567179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,12288,0.05602666404512194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,7168,0.040179553959104754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,6144,0.02626488937271966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,7168,0.0196088883611891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,7168,0.05197333296140035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,6144,0.017749332719379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,5120,0.024854222933451336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,6144,0.03547022077772353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,5120,0.015547555353906421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,8192,0.03784711162249247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,6144,0.04624799887339274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,4096,0.01997866729895274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,5120,0.031084444787767198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,3584,0.017893332574102614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,4096,0.013439111411571503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,5120,0.04030488928159078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,10240,0.06973510980606079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,4096,0.0258879992696974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,3072,0.015582222077581616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,3584,0.02312266661061181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,3584,0.012737777498033313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,2560,0.01386844449573093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,4096,0.03462400039037069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,3072,0.011787555283970304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,16384,0.03585244549645318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,3072,0.020956445071432326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,3584,0.031776888502968684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,2048,0.01200977795653873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,2560,0.01074488874938753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,2560,0.01797777745458815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,1536,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,2048,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,2560,0.025757332642873127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,2048,0.015517334143320719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,1024,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,1536,0.013122667041089801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,2048,0.021701332595613267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,1024,0.010484444598356882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,768,0.005640888793600931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,1536,0.01926399932967292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,8192,0.02110133402877384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,1024,0.016152888536453247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,512,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,768,0.009465778039561378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,768,0.015196444259749519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,512,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,256,0.0041751112374994485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,3072,0.028907554017172918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,512,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,512,0.013493333425786761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,128,0.0035573335157500375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,256,0.007276444799370236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,256,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,1536,0.008973333570692275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,64,0.003510222252872255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,128,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,128,0.007221333682537079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,32,0.003564444267087512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,1024,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,64,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,128,0.010057777994208867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,65536,0.18469866116841635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,32,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,768,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,51200,0.14475288656022814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,65536,0.1261777745352851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,16384,0.047188444270028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,65536,0.2866542339324951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,51200,0.10028888781865437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,16384,0.03435733252101474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,51200,0.2216035525004069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,12288,0.04206755426194933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,16384,0.07369777891370985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,10240,0.03701511025428772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,12288,0.027817779117160376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,65536,0.3834968937767877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,16384,0.10412355264027913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,51200,0.3007297780778673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,8192,0.02907022171550327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,10240,0.024731554918819006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,256,0.012464889221721225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,10240,0.04843644301096598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,7168,0.025608888930744592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,8192,0.020417778028382193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,8192,0.039854221873813205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,10240,0.06902666886647542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,6144,0.02288355595535702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,7168,0.03615466753641764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,7168,0.018823110395007663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,5120,0.018974221414989896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,8192,0.05716533131069607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,7168,0.051643557018703885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,6144,0.01704266667366028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,6144,0.031657778554492526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,4096,0.015697777271270752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,5120,0.02756799923049079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,6144,0.045144889089796275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,3584,0.013920888304710388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,4096,0.02290222214327918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,4096,0.013191111385822296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,12288,0.056511110729641385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,4096,0.03414400087462531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,3584,0.012424888710180918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,3584,0.02090577781200409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,3072,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,2560,0.010966222319338055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,12288,0.08061777883105807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,3072,0.011352889239788055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,3584,0.03102133340305752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,3072,0.018570666511853535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,2048,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,2560,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,2560,0.016194666425387066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,1536,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,3072,0.02815999918513828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,2048,0.009381333159075843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,2048,0.014135999812020196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,2560,0.02493155499299367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,1024,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,1536,0.008732444710201686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,1536,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,2048,0.021576000584496394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,5120,0.015150222513410779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,1024,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,1536,0.01871288816134135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,1024,0.009654222263230218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,512,0.004519111166397731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,5120,0.03996444410747952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,768,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,768,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,1024,0.015842666228612263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,256,0.003855111284388436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,512,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,512,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,768,0.014859555496109856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,256,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,256,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,512,0.013220444321632385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,64,0.003160888950030009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,256,0.01218488895230823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,128,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,32,0.0032035555276605818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,65536,0.19887999693552652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,128,0.0099973330895106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,64,0.006904888898134232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,32,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,768,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,51200,0.14412889215681288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,65536,0.08383733034133911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,65536,0.24647821320427787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,16384,0.05193600058555603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,51200,0.066575997405582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,65536,0.24343644248114693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,51200,0.19462133778466117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,16384,0.06349511279000176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,12288,0.0377511117193434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,128,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,16384,0.06982044378916423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,10240,0.028749333487616643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,12288,0.01962222158908844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,12288,0.04879200127389696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,12288,0.05622577667236328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,10240,0.017989334132936265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,8192,0.024432889289326135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,10240,0.04122489028506809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,10240,0.04831288920508491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,8192,0.014492443866199918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,8192,0.03491288754675124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,7168,0.021672000487645466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,51200,0.19064088662465414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,6144,0.01906133360332913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,128,0.0032248888164758682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,7168,0.013639111485746173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,8192,0.04068444503678216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,5120,0.016410667035314772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,16384,0.02346755564212799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,7168,0.03144355614980062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,4096,0.01367377738157908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,6144,0.027679999669392902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,6144,0.012432888978057437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,7168,0.03649333450529311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,5120,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,3584,0.012587555580668978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,4096,0.010191111101044549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,5120,0.02365066607793172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,6144,0.03285955720477634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,3072,0.011015999648306104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,4096,0.019759999381171334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,3584,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,3584,0.017902221944597032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,2560,0.010006222460005019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,4096,0.024477332830429077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,3072,0.009364444348547194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,3072,0.016376889414257474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,2048,0.00850844466023975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,3584,0.022680888573328655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,2560,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,1536,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,3072,0.020688888099458482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,2560,0.01443199978934394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,2048,0.008485332959228093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,1024,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,2048,0.01274577776590983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,2560,0.01887999971707662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,1536,0.007684444387753804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,768,0.0046746668716271715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,1536,0.01072355525361167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,2048,0.016551999582184684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,1024,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,512,0.004091555459631814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,1024,0.00903555585278405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,1536,0.014846222268210517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,768,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,768,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,256,0.0034924443397257063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,1024,0.013419555293189155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,512,0.00664533343580034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,128,0.002842666581273079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,512,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,768,0.012718222207493253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,256,0.006409777949253718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,64,0.002834666727317704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,256,0.006654222392373615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,5120,0.029027554723951552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,32,0.002833777831660377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,512,0.011796444654464722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,128,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,128,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,256,0.01071466671095954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,65536,0.1583591169781155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,64,0.006510222123728857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,32,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,128,0.008873778084913889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,51200,0.12370133399963379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,16384,0.041548444165123835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,65536,0.07938133345709907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,65536,0.20031466748979357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,51200,0.06253955761591594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,12288,0.030279109875361126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,16384,0.022950222094853718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,16384,0.058654222223493785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,65536,0.23863111601935494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,10240,0.025407110651334126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,51200,0.1915644407272339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,12288,0.021137777302000258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,12288,0.04048711061477661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,16384,0.06868533293406169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,8192,0.020029332902696397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,10240,0.017509332961506315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,10240,0.03497155507405599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,7168,0.018591110905011494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,8192,0.02944799926545885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,8192,0.014139556222491793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,10240,0.04696444339222378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,6144,0.015487111277050443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,8192,0.03939466675122579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,7168,0.013094222380055321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,7168,0.026587555805842083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,5120,0.01368799971209632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,6144,0.012106666962305704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,6144,0.023783110909991797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,7168,0.03597955571280585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,4096,0.011585777832402123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,5120,0.020208888583713107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,51200,0.15847199492984348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,5120,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,6144,0.03238222334120009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,3584,0.010664000279373594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,4096,0.01000533335738712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,5120,0.028299556838141546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,3072,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,3584,0.009665778113736046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,3584,0.01573422220018175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,4096,0.023997333314683702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,2560,0.008361777497662438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,12288,0.054273777537875705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,3584,0.022336888644430373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,3072,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,2048,0.007252444823582967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,2560,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,2560,0.012488889197508493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,3072,0.02016266683737437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,1536,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,2048,0.008045333127180735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,2048,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,2560,0.0185244447655148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,1024,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,1536,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,2048,0.016543999314308167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,768,0.0042168887125121225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,1024,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,1024,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,1536,0.01479466590616438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,512,0.0036844445599450003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,1024,0.01273155543539259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,768,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,4096,0.017218665944205392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,256,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,512,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,512,0.006775110960006714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,768,0.012031111452314587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,3072,0.009007111191749573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,128,0.0031537777847713898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,256,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,256,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,64,0.00281866660548581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,128,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,32,0.002993777808215883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,256,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,128,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,65536,0.10228621959686279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,128,0.008352888955010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,1536,0.00965244405799442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,32,0.00620977787507905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,51200,0.07147377729415894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,65536,0.1557368834813436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,768,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,16384,0.03220444586541917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,51200,0.06263733572430082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,51200,0.1221875614590115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,65536,0.22835999064975313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,12288,0.024254222710927326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,16384,0.02197955548763275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,16384,0.04219111137919956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,51200,0.18104088306427002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,10240,0.020733333296246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,512,0.011027555498811932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,12288,0.033077332708570696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,16384,0.06523378027809991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,12288,0.018577777677112155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,8192,0.01699022286468082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,10240,0.0166657782263226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,12288,0.05108799868159824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,64,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,10240,0.028591109646691218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,7168,0.015454222758611044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,8192,0.023875556058353845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,8192,0.0136808885468377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,10240,0.04433511031998528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,65536,0.0781608886188931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,7168,0.02165422174665663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,7168,0.012799999780125089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,5120,0.01185066666867998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,8192,0.036770668294694685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,6144,0.011711999773979187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,7168,0.03400355577468872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,6144,0.019223110543357003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,4096,0.009988444546858469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,5120,0.011012444065676795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,6144,0.02978133492999607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,3584,0.009134222235944536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,4096,0.014163555370436775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,4096,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,5120,0.026520889666345384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,3072,0.008143111235565608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,3584,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,3584,0.013101333545313941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,3072,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,3072,0.01204533295498954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,2560,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,3584,0.02110311057832506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,6144,0.01367377738157908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,2048,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,2560,0.008683555656009251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,3072,0.01924711134698656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,2560,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,1536,0.005392000079154968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,2048,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,2560,0.017887999614079792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,2048,0.009710222482681274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,1024,0.004518222063779831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,5120,0.01661422186427646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,1536,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,2048,0.01570044457912445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,768,0.0038737778862317405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,1024,0.007233777807818518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,1024,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,1536,0.014140443669425117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,512,0.003220444545149803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,4096,0.023037332627508376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,768,0.006666666517655055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,768,0.006909333169460297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,256,0.0028826666788922418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,512,0.006599111275540457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,512,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,128,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,768,0.012082666986518435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,256,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,256,0.005879999862776862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,512,0.011297778122954898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,64,0.002471999989615546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,32,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,128,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,256,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,128,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,65536,0.08950577841864692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,64,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,32,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,128,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,51200,0.07043466965357463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,1536,0.008421333299742805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,65536,0.07728266716003418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,65536,0.1514257722430759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,51200,0.06265866756439209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,51200,0.1202053361468845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,65536,0.2255208757188585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,12288,0.01902666687965393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,16384,0.04098666707674662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,1024,0.012798222402731577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,51200,0.1791013346778022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,10240,0.01610311037964291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,12288,0.018221333622932434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,16384,0.06453688939412434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,8192,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,12288,0.03257955445183648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,10240,0.016414221790101793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,10240,0.028083556228213843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,12288,0.050923556089401245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,7168,0.012364444633324942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,8192,0.013639999760521783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,8192,0.023518222901556227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,10240,0.04361777835422092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,6144,0.011038222246699862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,7168,0.012721777790122561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,7168,0.02104533380932278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,8192,0.03657333387268914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,5120,0.009837333526876237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,16384,0.024653333756658766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,7168,0.0335457788573371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,6144,0.029153777493370905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,5120,0.016363556186358135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,4096,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,5120,0.01069777790043089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,16384,0.02164977788925171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,3584,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,4096,0.009865778187910715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,5120,0.026201777988009985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,4096,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,3072,0.006724444528420766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,3584,0.012793777717484368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,4096,0.022376888328128394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,3072,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,3072,0.011756444142924415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,2560,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,3072,0.019162666466501024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,2560,0.010671111444632212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,6144,0.019096000327004325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,2048,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,2560,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,2560,0.017575111654069688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,1536,0.004625777817434735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,2048,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,2048,0.009504000345865885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,1024,0.003846222327815162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,2048,0.015829333000712924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,1536,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,3584,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,1024,0.00703022215101454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,1536,0.014099554883109199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,1024,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,3584,0.02094488839308421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,1024,0.012426666915416718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,768,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,768,0.011348444554540845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,6144,0.011757333245542316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,512,0.006173333360089197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,256,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,512,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,512,0.01108088923825158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,128,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,256,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,256,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,1536,0.008336000144481659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,256,0.010701333483060202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,64,0.002513777878549364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,128,0.005494222044944763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,768,0.003527111063400904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,128,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,65536,0.05606666538450453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,64,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,32,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,768,0.0068488890926043195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,51200,0.04566133353445265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,65536,0.07793955670462714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,65536,0.11084800296359593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,16384,0.01755199995305803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,51200,0.06247466802597046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,51200,0.08734933535257976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,65536,0.224397341410319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,16384,0.02075466679202186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,16384,0.031136890252431233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,51200,0.17810932795206705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,12288,0.017627555463049147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,128,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,16384,0.06408266888724433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,32,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,12288,0.024875554773542616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,8192,0.0129466669427024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,10240,0.016343111793200176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,12288,0.04980088935958015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,10240,0.021320889393488567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,7168,0.011683555940786997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,512,0.002867555452717675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,8192,0.017937777770890128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,8192,0.013455111119482251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,6144,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,10240,0.04309511184692383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,7168,0.012439111040698158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,7168,0.016923555069499545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,8192,0.03616888986693488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,5120,0.009004444711738164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,6144,0.011540444360838996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,7168,0.03288977675967746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,4096,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,12288,0.017009778155220878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,5120,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,6144,0.028965334097544353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,5120,0.013453333742088742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,3584,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,4096,0.009678222239017487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,5120,0.025663110944959853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,4096,0.011559111376603445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,3072,0.006227555374304454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,10240,0.015612444943851896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,3584,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,3584,0.009186666872766282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,2560,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,4096,0.022128888302379187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,3584,0.020925333102544148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,3072,0.009999111294746399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,2048,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,2560,0.00906577789121204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,2560,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,2048,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,2048,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,1536,0.0041884444653987885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,6144,0.014895111322402954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,2048,0.015166223049163818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,1024,0.0035546666218174826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,1536,0.007484444313579136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,1536,0.014000000225173103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,1024,0.006335111127959357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,768,0.0031715554909573663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,1024,0.006945778098371293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,3072,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,1024,0.012039999994966718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,512,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,768,0.006693333387374878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,768,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,3072,0.018935999936527677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,256,0.0025111111915773815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,512,0.00645955569214291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,768,0.011701333026091257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,128,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,2560,0.017542221479945712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,256,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,256,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,64,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,512,0.010448888772063784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,32,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,128,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,128,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,1536,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,65536,0.03672622309790717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,256,0.009665778113736046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,51200,0.02929511004024082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,128,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,64,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,32,0.0052933332820733385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,65536,0.07697066995832655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,16384,0.012991999586423239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,65536,0.10908710956573486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,51200,0.061779558658599854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,12288,0.010869333313571082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,51200,0.08616089158587986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,16384,0.02070399953259362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,16384,0.03013777732849121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,65536,0.22192888789706758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,10240,0.011426666544543372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,51200,0.17782044410705566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,12288,0.01761866609255473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,16384,0.06338489055633545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,8192,0.01238222254647149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,12288,0.02384533319208357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,512,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,7168,0.011060444845093621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,10240,0.0162444445821974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,10240,0.02086755633354187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,8192,0.013460444079505073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,6144,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,8192,0.017678222722477384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,7168,0.01239466667175293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,10240,0.041856888267729014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,7168,0.016207999653286405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,5120,0.008347555167145198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,8192,0.03494666682349311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,6144,0.011417778001891242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,6144,0.014765333798196582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,4096,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,7168,0.031896889209747314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,5120,0.010690666735172272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,5120,0.013071999781661563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,3584,0.006595555692911148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,6144,0.028547555208206177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,4096,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,5120,0.025430222352345783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,3072,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,3584,0.009200000100665623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,3584,0.010646222366227044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,2560,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,3072,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,3072,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,3072,0.01884888940387302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,2560,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,2560,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,2048,0.004546666724814309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,2048,0.00775022225247489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,2560,0.017229333519935608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,1536,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,2048,0.008117333054542542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,2048,0.015145777000321282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,1536,0.007344000041484833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,1024,0.0031964443624019623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,1536,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,4096,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,1024,0.006994666324721442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,768,0.003031999907559819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,1024,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,4096,0.02172800070709652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,1024,0.01204622205760744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,512,0.0025635556214385563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,3584,0.020614221692085266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,768,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,768,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,256,0.002428444516327646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,512,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,12288,0.04883822136455112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,768,0.011109333071443768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,128,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,256,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,256,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,512,0.011029333704047732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,64,0.0018542222678661346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,128,0.005130666825506422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,128,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,32,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,256,0.009672000176376766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,65536,0.02966844373279148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,64,0.004960000101062986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,128,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,32,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,51200,0.0239991115199195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,1536,0.014123555686738757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,65536,0.10834133625030518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,65536,0.07672533061769274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,16384,0.010227555202113258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,51200,0.061556445227728955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,51200,0.08517155382368301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,65536,0.22385244899325898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,16384,0.020655110478401184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,12288,0.011759999725553723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,16384,0.02973066767056783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,10240,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,51200,0.1779493358400133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,512,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,12288,0.017500445246696472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,8192,0.009092444347010719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,16384,0.06220710939831204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,12288,0.023588443795839947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,10240,0.01586844523747762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,10240,0.02057066725360023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,12288,0.04819555415047539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,8192,0.017522666189405654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,6144,0.008015111088752747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,10240,0.04145955708291795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,7168,0.016229333149062265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,8192,0.034866667456097074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,5120,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,6144,0.01145244472556644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,6144,0.014554666148291694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,7168,0.03190044562021891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,4096,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,5120,0.010363555616802638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,6144,0.02847377790345086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,5120,0.012969777815871768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,3584,0.006211555666393704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,4096,0.011372444530328115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,5120,0.025073778298166063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,3584,0.010386666489972008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,3584,0.009083555804358589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,4096,0.021935110290845234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,7168,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,8192,0.01311555587583118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,2560,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,3584,0.020280889338917203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,3072,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,7168,0.012455999851226807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,2048,0.004545777622196409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,2560,0.008924444516499838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,2560,0.00835111074977451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,3072,0.018352000249756705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,1536,0.0038968887594011095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,2048,0.007690666450394525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,2560,0.01718133356836107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,2048,0.008343111309740279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,1024,0.0031902222997612427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,1536,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,1536,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,2048,0.014826666977670459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,4096,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,768,0.0028773333049482773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,1024,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,1536,0.013778666655222574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,3072,0.005535111245181825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,768,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,768,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,256,0.002448888909485605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,512,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,512,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,256,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,256,0.005966222120655908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,128,0.002130666747689247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,512,0.01072177787621816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,128,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,256,0.010048888623714447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,64,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,128,0.004948444250557157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,32,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,128,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,65536,0.029080000188615587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,64,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,32,0.004978666702906291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,51200,0.022703111171722412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,16384,0.009122666385438707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,512,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,12288,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,1024,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,65536,0.10820800065994263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,10240,0.007461333440409766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,51200,0.08499911096360947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,8192,0.007254222200976477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,16384,0.029273778200149536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,3072,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,7168,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,6144,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,12288,0.023358222511079576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,5120,0.006657777975002925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,4096,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,10240,0.020385776956876118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,3584,0.006218666831652324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,8192,0.01750666730933719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,7168,0.015839111473825242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,3072,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,6144,0.014448889427714877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,2560,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,5120,0.012807110945383707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,2048,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,1536,0.0037164443896876443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,3584,0.010395555860466426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,3072,0.009646221995353699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,1024,0.003162666741344664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,2560,0.00869777798652649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,768,0.00282577777074443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,2048,0.008080888953473832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,512,0.0024844445288181305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,256,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,1536,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,128,0.001956444440616502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,1024,0.006228444476922353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,768,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,64,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,512,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,256,0.005048888838953442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,65536,0.02569955587387085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,128,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,51200,0.02027466727627648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,16384,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,768,0.011071999867757162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,64,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,32,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,10240,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,65536,0.10822044478522407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,51200,0.08512088987562393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,8192,0.007637333538797166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,16384,0.029208001163270738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,7168,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,6144,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,12288,0.023307555251651343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,5120,0.005990222096443176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,10240,0.02032888929049174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,4096,0.011085333095656501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,4096,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,3584,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,8192,0.017364444004164804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,3072,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,7168,0.01588000026014116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,2560,0.004860444615284602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,2048,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,5120,0.01278133359220293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,4096,0.01129422254032559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,1536,0.004200889004601372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,3584,0.01036711119943195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,1024,0.0034933334423436057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,3072,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,768,0.00300799993177255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,2560,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,2048,0.008053333395057255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,256,0.002151111140847206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,1536,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,1024,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,128,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,64,0.0018728888697094387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,512,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,12288,0.008977777428097194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,256,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,128,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,64,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,32,0.004872888740566042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,6144,0.01442311041884952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,512,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,768,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,16384,0.4017537699805365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,16384,1.1355821821424696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,12288,0.7207244237263998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,12288,0.3039066791534424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,12288,0.8154506683349609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,10240,0.6325226889716254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,16384,0.9906329049004449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,51200,1.4091786278618708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,10240,0.2570142216152615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,8192,0.504095130496555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,10240,0.7029600143432617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,51200,3.191859563191732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,8192,0.5376924408806695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,7168,0.4519342316521539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,8192,0.2234222094217936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,7168,0.18593245082431367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,6144,0.415812439388699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,5120,0.3620595667097304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,6144,0.15964444478352866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,7168,0.482044431898329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,6144,0.4008951187133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,51200,3.7990086873372397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,4096,0.29960444238450795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,5120,0.32943201065063477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,3584,0.2085191143883599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,4096,0.11492889457278782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,4096,0.2749137878417969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,3584,0.11578133371141221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,5120,0.14025955730014378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,3072,0.20127556059095594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,3584,0.23094132211473253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,2560,0.15105156103769937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,3072,0.08974755472607082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,2048,0.13671822018093532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,2560,0.07852800024880303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,1536,0.10604888863033718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,3072,0.20027732849121094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,2560,0.17863467004564074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,2048,0.1400782267252604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,768,0.057061334451039634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,1536,0.05254044466548496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,1536,0.11341244644588894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,512,0.04575466778543261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,1024,0.08418755398856269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,768,0.038386666112475924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,256,0.029666665527555678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,512,0.035432888401879206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,128,0.02338933282428318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,512,0.060030221939086914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,768,0.06805155674616496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,256,0.032836443848080106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,64,0.020097777247428894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,256,0.05398577782842848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,32,0.0206675562593672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,128,0.032016889916525945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,128,0.05027022295527988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,64,0.049978666835361056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,32,0.049954666031731494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,2048,0.06655644708209567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,1024,0.04079644547568427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,16384,0.7671279907226562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,1024,0.08271733257505628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,16384,0.3193875683678521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,51200,1.0690444310506184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,65536,1.3898169199625652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,16384,0.9034542507595487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,51200,2.606620364718967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,10240,0.45512088139851886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,12288,0.714445326063368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,51200,2.9225234985351562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,65536,3.385598288642036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,12288,0.2708666589525011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,12288,0.6830248832702637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,8192,0.44898043738471133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,10240,0.19967910978529188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,65536,3.634688059488932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,7168,0.3937377664777968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,10240,0.5360426902770996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,8192,0.4084542327457004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,6144,0.28215111626519096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,5120,0.231316434012519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,8192,0.16651999950408936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,7168,0.3686231242285834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,7168,0.15294133292304143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,6144,0.14717333846622044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,4096,0.1848897801505195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,6144,0.31302934222751194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,5120,0.25908978780110675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,3584,0.19546843899620903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,3072,0.1430248949262831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,4096,0.21039022339714897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,3584,0.08335822158389622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,2560,0.12938577598995632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,3584,0.24981954362657335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,3072,0.0771991080707974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,2048,0.09778666496276855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,2560,0.06191821893056234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,2560,0.13483199808332655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,3072,0.15736356046464708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,2048,0.05423644516203138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,1024,0.05639555719163683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,2048,0.11797866556379531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,1536,0.04265244470702278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,1536,0.08955199850930108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,1024,0.03306577934159173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,768,0.05112888746791416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,1024,0.06998844279183282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,512,0.0314977765083313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,768,0.030990223089853924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,256,0.02383288906680213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,768,0.05489333470662435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,512,0.028959999481836956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,4096,0.08990577856699626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,5120,0.11005955272250706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,256,0.026939556002616882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,256,0.04341333442264133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,64,0.01791733337773217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,32,0.018955555227067735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,128,0.040431999497943454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,64,0.04003466831313239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,32,0.04011022382312351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,1536,0.07792888747321235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,128,0.021553777986102637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,512,0.047803554270002574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,65536,0.9886444939507378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,128,0.026259554757012263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,16384,0.24055377642313638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,51200,0.7891608874003092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,12288,0.1822133329179552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,16384,0.10804888937208389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,65536,1.1541431215074327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,51200,0.3559546735551622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,16384,0.3112000094519721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,10240,0.16328622235192194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,12288,0.21270667182074654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,12288,0.08415466547012329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,8192,0.14726932843526205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,10240,0.07907644245359632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,10240,0.17256711588965523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,7168,0.12096622255113389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,51200,0.9014346864488391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,8192,0.057368887795342334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,6144,0.10656533638636272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,8192,0.1389875544442071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,7168,0.05363022287686666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,5120,0.10108977556228638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,7168,0.12370755937364365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,6144,0.04584622383117676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,4096,0.07814133167266846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,5120,0.039965334865782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,6144,0.10551999674903022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,3584,0.06135644515355428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,4096,0.03286577926741706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,65536,0.4428399933709039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,5120,0.0884551140997145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,4096,0.07304444577958849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,3584,0.029273778200149536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,2560,0.05403022302521599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,3584,0.06570133235719469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,2048,0.04470844401253594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,3072,0.02632177703910404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,1536,0.03454577922821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,2560,0.023927999867333308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,3072,0.05648889144261678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,1024,0.023111111587948267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,2560,0.04871911141607496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,2048,0.020796444680955675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,768,0.018256889449225534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,2048,0.04055733482042948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,512,0.013855111267831592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,1024,0.015096000499195524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,768,0.013665777941544851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,1024,0.025391111771265667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,256,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,768,0.0215022216240565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,512,0.01242844429281023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,128,0.008801778157552084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,512,0.017909333109855652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,256,0.011731555892361535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,64,0.00812355594502555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,32,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,128,0.011577777564525604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,128,0.015558222929636637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,64,0.015486222174432544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,3072,0.05654044283760918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,32,0.015451555450757345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,1536,0.01776711146036784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,1536,0.03281689021322463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,51200,0.5777075555589464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,65536,0.3440408971574571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,16384,0.18266933494144014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,65536,0.8781955507066516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,16384,0.08687111404207017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,51200,0.6887813144259982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,12288,0.1453146669599745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,256,0.015811555915408664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,51200,0.27087555991278756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,10240,0.12218933635287815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,16384,0.2239875528547499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,12288,0.06704177459081014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,8192,0.11307644844055176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,10240,0.05773244301478068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,12288,0.15540533595614964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,7168,0.0998106665081448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,8192,0.04658577839533488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,10240,0.13169599903954401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,6144,0.08527288834253947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,8192,0.10589066478941177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,5120,0.07305777735180326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,7168,0.09746133618884617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,6144,0.036962668100992836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,4096,0.057236446274651416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,6144,0.0821031133333842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,5120,0.03610488772392273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,3584,0.04604533314704895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,65536,0.8152329126993815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,5120,0.06934577888912626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,4096,0.02762311034732395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,3072,0.04313866628540886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,4096,0.05771911144256592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,3584,0.025111999776628282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,2560,0.037984000311957464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,3072,0.02277511192692651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,3584,0.05350755651791891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,2048,0.031297776434156634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,1536,0.02392622166209751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,2048,0.017986666825082567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,2048,0.03245333168241713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,1536,0.02603111167748769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,1536,0.015484443969196744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,1024,0.016406221522225272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,768,0.013568888935777875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,1024,0.01294311136007309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,7168,0.04136711027887132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,1024,0.020663999848895602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,512,0.011056889262464313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,768,0.01757244434621599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,768,0.012096888489193387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,256,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,512,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,128,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,256,0.010510221951537661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,256,0.013351110948456658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,64,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,128,0.012788444757461548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,2560,0.020260444945759244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,32,0.006551110910044775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,2560,0.038577778471840754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,64,0.012798222402731577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,32,0.012795555922720166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,65536,0.6014097531636556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,65536,0.2934888998667399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,51200,0.4959431224399143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,512,0.01444533301724328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,16384,0.17557511064741346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,65536,0.7408426602681478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,128,0.010514666636784872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,51200,0.5929733382331001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,12288,0.13820266723632812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,16384,0.07445422146055433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,3072,0.04424355427424113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,10240,0.1039280030462477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,12288,0.05695111221737332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,8192,0.08899822500016953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,10240,0.04923999971813626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,12288,0.1329546637005276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,10240,0.12048266993628608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,8192,0.03987288806173537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,6144,0.07000711229112414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,8192,0.09108444717195298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,7168,0.035976890060636736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,5120,0.05725955300860935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,7168,0.0801795588599311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,6144,0.03257155418395996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,4096,0.04979377653863695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,6144,0.07051022185219659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,5120,0.028177777926127117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,3584,0.043306665288077466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,5120,0.06013333135181003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,3072,0.033111111985312566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,3584,0.021910222040282354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,4096,0.04924266537030538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,2560,0.030232889784706965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,3584,0.04469066527154711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,3072,0.01998488936159346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,2048,0.025384000606007043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,3072,0.03905866543451945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,2560,0.01812177730931176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,51200,0.23425777753194174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,1536,0.019465777609083388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,2560,0.0340320004357232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,1024,0.013537777794731988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,2048,0.016213332613309223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,7168,0.07031466563542683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,1536,0.013832888669437833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,2048,0.028913779391182795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,1024,0.01202755586968528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,512,0.00869688888390859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,768,0.011076444553004371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,768,0.015469332536061605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,512,0.010060444474220276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,256,0.00620177760720253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,512,0.013073777986897362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,4096,0.023711999257405598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,128,0.005656888915432825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,256,0.009683555199040307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,256,0.012094222009181976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,64,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,128,0.009679111341635386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,32,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,128,0.011703111231327057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,16384,0.19239732954237196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,64,0.011770666473441653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,32,0.011547555526097616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,768,0.011071999867757162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,1536,0.023166222704781428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,1024,0.01870044403605991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,65536,0.5374124315049913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,65536,0.2361537880367703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,16384,0.14600088861253527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,51200,0.36763911777072483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,12288,0.11392088731129964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,51200,0.1879955530166626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,16384,0.06100710895326403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,65536,0.5971466700236002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,10240,0.09212888611687554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,51200,0.47087377972073025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,12288,0.04747466577423943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,8192,0.07236978080537584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,10240,0.040961778826183744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,12288,0.1120835542678833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,7168,0.06457244687610202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,10240,0.09419911437564427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,8192,0.03302399979697333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,6144,0.0510044429037306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,7168,0.030677331818474665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,8192,0.07597333192825317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,5120,0.04339289002948337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,7168,0.06744888756010267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,4096,0.03750933210055033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,6144,0.026926222774717543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,3584,0.034316443734698825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,5120,0.024409777588314478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,6144,0.05935821930567423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,3072,0.03038133184115092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,4096,0.020616888999938965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,3584,0.019154666198624503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,4096,0.04139822059207492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,3072,0.017490666773584154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,2560,0.024160888459947374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,2048,0.020796444680955675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,3072,0.03313422203063965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,2560,0.016122667325867545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,1536,0.017310221989949543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,2560,0.029001779026455347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,2048,0.014363555444611443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,1024,0.013070222404268054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,16384,0.15302844842274985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,768,0.010854221880435944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,1536,0.012310222619109683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,1536,0.020043555233213637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,1024,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,512,0.008058666355080074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,1024,0.015621332658661736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,768,0.010038221875826517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,256,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,768,0.013417777915795645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,128,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,512,0.009540444446934594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,256,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,512,0.011432888607184092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,64,0.005007110950019625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,128,0.008958222137557136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,256,0.010464000205198923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,32,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,5120,0.05042577783266703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,128,0.010072000324726105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,64,0.01036622209681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,3584,0.03761511047681173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,32,0.010060444474220276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,65536,0.3932995531294081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,16384,0.10427555772993301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,51200,0.3709004455142551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,65536,0.22274221314324272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,12288,0.08584888776143391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,51200,0.17710933420393202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,2048,0.02459822263982561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,65536,0.532842689090305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,10240,0.08248977528678046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,51200,0.41010133425394696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,12288,0.04580444428655836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,8192,0.06791466474533081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,12288,0.09734133217069839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,10240,0.03879466652870178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,7168,0.06071644359164768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,6144,0.049272000789642334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,8192,0.032630221711264715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,7168,0.029324445459577773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,8192,0.06680355469385783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,5120,0.0417866673734453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,7168,0.05995821952819824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,6144,0.026526222626368206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,4096,0.032441778315438166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,6144,0.05390310949749417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,5120,0.02351288828584883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,3584,0.02939644455909729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,5120,0.04395644532309639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,4096,0.020942222740915086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,4096,0.036770668294694685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,3072,0.02511466708448198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,3584,0.01835911141501533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,2560,0.021320889393488567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,3584,0.033189333147472806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,3072,0.016916443904240925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,2048,0.017356443736288283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,16384,0.05725955300860935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,3072,0.02947288751602173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,2560,0.015111999379263984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,1536,0.01349422252840466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,16384,0.1276479959487915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,2048,0.013645333548386892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,1024,0.010094222095277574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,1536,0.012082666986518435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,768,0.008264888491895463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,10240,0.08195110824373034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,1536,0.017992888887723286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,1024,0.010195555786291758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,512,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,1024,0.014096889230940076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,768,0.009391110804345872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,256,0.005208000126812193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,512,0.008850666383902231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,768,0.012435555458068848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,128,0.004579555657174853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,256,0.0086666668454806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,64,0.0041857775714662345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,128,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,32,0.004252444538805219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,128,0.009679999616410997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,64,0.009650666680600908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,32,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,65536,0.35367290178934735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,2560,0.02570044497648875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,2048,0.022076444493399724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,65536,0.18410310480329725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,51200,0.2770773304833306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,16384,0.11002844572067261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,51200,0.1561679972542657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,512,0.010623111493057676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,51200,0.3523057831658258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,12288,0.08236089017656115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,16384,0.0496693319744534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,256,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,10240,0.0716186695628696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,16384,0.11829600069257949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,8192,0.05459288756052653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,12288,0.04101599918471442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,10240,0.0337759984864129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,7168,0.0442755553457472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,8192,0.027709333433045283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,6144,0.037074665228525795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,12288,0.08561688661575317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,7168,0.025081778566042583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,5120,0.03527377711402045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,6144,0.02277244461907281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,4096,0.02747111188040839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,5120,0.020081778367360432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,6144,0.04713955521583557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,3584,0.024480889240900677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,5120,0.039691554175482854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,4096,0.018242667118708294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,3072,0.021520889467663232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,4096,0.03373422225316366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,65536,0.46688800387912327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,3584,0.030392888519499037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,2560,0.018618666463428073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,3072,0.015071999695565967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,2048,0.015472888946533203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,3072,0.027162666122118633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,2560,0.013827555709415011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,1536,0.012544000314341651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,2560,0.023723555935753718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,2048,0.012787555654843649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,1024,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,2048,0.020403555697864957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,1536,0.011343999869293638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,10240,0.07391022311316596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,768,0.008184889124499427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,1536,0.016600888636377122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,8192,0.0595911078982883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,1024,0.009661333428488838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,7168,0.05342400074005127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,768,0.009004444711738164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,1024,0.012940444052219391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,256,0.005012444323963589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,512,0.008600888980759515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,128,0.004197333421972063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,768,0.011424889167149862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,64,0.004269333349333869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,256,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,32,0.004477333277463913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,256,0.009186666872766282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,3584,0.016269332832760282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,128,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,64,0.00868977771864997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,32,0.008728889127572378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,65536,0.3827422195010715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,65536,0.17121067312028673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,51200,0.2574133343166775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,65536,0.40367645687527126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,16384,0.0834862192471822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,512,0.00656088896923595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,51200,0.13594933350880942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,12288,0.06463555494944255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,16384,0.046923554605907865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,51200,0.3492364353603787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,10240,0.05928444200091892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,16384,0.099016891585456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,512,0.009748444788985783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,12288,0.03706399930848016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,8192,0.039934221241209246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,128,0.008027555214034187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,7168,0.03502488798565335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,12288,0.07657333215077718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,10240,0.0321377780702379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,8192,0.02623733381430308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,10240,0.06618755393558078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,5120,0.02997066577275594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,7168,0.024028443627887305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,8192,0.053232000933753125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,6144,0.02164266672399309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,4096,0.02416177756256527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,6144,0.04133866561783685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,5120,0.019792889555295307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,3584,0.02163377735349867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,4096,0.016907556189431083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,5120,0.03626755542225308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,3072,0.019395555059115093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,4096,0.030243555704752605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,3584,0.015426667200194465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,2560,0.0165955556763543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,3584,0.027639999985694885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,3072,0.0140684445699056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,2048,0.013855111267831592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,3072,0.024199111594094172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,2560,0.012992888689041138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,1536,0.011352889239788055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,2560,0.021526222427686054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,1024,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,2048,0.018223111828168232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,1536,0.010738666686746808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,768,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,1536,0.014811555544535318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,1024,0.00942844483587477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,512,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,6144,0.03511910968356662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,1024,0.01180266671710544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,768,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,256,0.004508444418509801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,768,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,128,0.003875555677546395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,512,0.008380444513426887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,7168,0.04733777708477444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,64,0.0035626664757728577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,256,0.007867555651399825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,128,0.007930667036109501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,256,0.008619555168681675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,128,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,64,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,32,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,65536,0.26678310500250924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,65536,0.13122844696044922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,51200,0.2112435499827067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,2048,0.0120319997270902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,16384,0.0764568911658393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,65536,0.33127466837565106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,12288,0.060228447119394936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,51200,0.10343288713031346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,10240,0.04872622092564901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,51200,0.25586043463812935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,16384,0.08178400331073336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,12288,0.06300355328453912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,8192,0.038035555018319026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,10240,0.025290666355027094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,7168,0.03581066595183478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,10240,0.05364888906478882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,8192,0.020862221717834473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,512,0.009031111167536842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,6144,0.02828533450762431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,8192,0.04424000117513868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,7168,0.019300444258583915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,5120,0.02373955481582218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,6144,0.017587555779351127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,7168,0.03933866818745931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,4096,0.020078221956888836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,6144,0.03476000163290236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,5120,0.01629688839117686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,3584,0.018263111511866253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,5120,0.03015111221207513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,4096,0.013799111048380533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,3072,0.016350222958458793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,4096,0.02535466684235467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,3584,0.023189332750108507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,2560,0.01371911085314221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,3072,0.020711110697852243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,16384,0.03683822353680929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,12288,0.02850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,2560,0.010758221977286868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,2560,0.018238221605618794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,2048,0.009655111365848118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,2048,0.015579554769727917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,1536,0.009950222240553962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,1024,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,1536,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,1536,0.0127368892232577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,768,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,1024,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,768,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,3584,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,512,0.0052968888647026485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,3072,0.011868444581826528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,256,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,512,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,512,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,2048,0.012105777859687805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,128,0.0038542221817705366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,256,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,64,0.0035546666218174826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,256,0.00796977761718962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,32,0.003564444267087512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,32,0.003540444291300244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,128,0.007677333222495184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,64,0.007667555577225155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,32,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,65536,0.26718489329020184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,1024,0.010463111102581024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,51200,0.17647200160556367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,65536,0.12494666046566433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,65536,0.28997333844502765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,16384,0.05577422512902153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,12288,0.04114400015936957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,768,0.009384888741705153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,51200,0.1006044414308336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,16384,0.036044445302751325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,51200,0.22619289822048613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,10240,0.04140355520778232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,16384,0.07292711072497897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,8192,0.03303022185961405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,12288,0.028389334678649902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,7168,0.028534223635991413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,12288,0.057562669118245445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,6144,0.022712000542216833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,8192,0.02093688812520769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,10240,0.047480000389946826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,5120,0.020689777202076383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,128,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,7168,0.019111111760139465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,4096,0.01713511182202233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,7168,0.03549422158135308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,6144,0.01734755602147844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,3584,0.014920888675583733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,6144,0.031210667557186548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,5120,0.026901332868470088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,3072,0.013777777552604675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,3584,0.012730666332774691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,3584,0.020990222692489624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,2560,0.011570666399266986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,3072,0.01187999960449007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,3072,0.018914666440751817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,2048,0.010336000058386061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,2560,0.0110880004035102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,2560,0.015929778416951496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,1536,0.008291555775536431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,2048,0.00996977753109402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,2048,0.013921777407328287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,10240,0.025231111380789015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,1024,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,1536,0.009731555150614845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,768,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,8192,0.03886311252911886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,1024,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,512,0.004534222185611725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,1024,0.009710222482681274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,768,0.007924444145626491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,256,0.003597333199448056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,768,0.008672888908121321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,512,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,512,0.007790221936172909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,4096,0.022830223043759663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,256,0.007069333559936947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,64,0.003020444470975134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,256,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,32,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,128,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,128,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,64,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,65536,0.23799734645419648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,51200,0.18657422065734863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,65536,0.1217413345972697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,65536,0.2702551152971056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,1536,0.011639110743999481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,16384,0.06281066603130765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,5120,0.0162426663769616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,51200,0.09729511208004421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,4096,0.013712888790501488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,16384,0.06926933262083265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,51200,0.20749955707126191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,16384,0.03456977672047085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,10240,0.041165331999460854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,12288,0.027797334724002417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,8192,0.03418488966094123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,10240,0.02478222217824724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,10240,0.045447998576694064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,7168,0.028660443094041612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,8192,0.020389333367347717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,32,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,6144,0.025430222352345783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,7168,0.018584888842370775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,7168,0.03403288788265652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,5120,0.02141777839925554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,6144,0.016516443755891588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,6144,0.029923554923799302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,4096,0.01779022150569492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,5120,0.01499022212293413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,5120,0.025977777110205755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,12288,0.050428443484836154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,3584,0.015840000576443143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,128,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,4096,0.012986666626400419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,3072,0.014139556222491793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,3584,0.012553777959611682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,2560,0.012437333663304647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,3584,0.020111999577946134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,12288,0.05361244413587782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,2048,0.010476444330480365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,3072,0.017779555585649278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,2560,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,1536,0.008422222402360704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,8192,0.037709332174725003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,2048,0.013765333427323235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,1024,0.006505777852402792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,1536,0.011415111521879831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,1536,0.008761777646011775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,768,0.005499555418888728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,1024,0.008375110725561777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,1024,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,512,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,768,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,768,0.008643555144468943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,256,0.00424177779091729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,512,0.007425777614116669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,4096,0.02216088937388526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,128,0.0034826666944556763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,256,0.007068444457319047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,512,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,64,0.003212444484233856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,3072,0.011434666812419891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,32,0.003549333247873518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,128,0.006989333364698622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,128,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,65536,0.16941955354478624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,64,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,2560,0.01556622154182858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,32,0.0069848886794514125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,2048,0.009574222067991892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,51200,0.1300897730721368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,16384,0.04855111241340637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,12288,0.03662933243645562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,65536,0.1196062167485555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,51200,0.0951111117998759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,65536,0.23025332556830513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,10240,0.02980088856485155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,16384,0.03309066759215461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,51200,0.1792142258750068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,8192,0.023914666639433965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,12288,0.026574222577942744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,16384,0.059522668520609535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,7168,0.02108977735042572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,10240,0.023689778314696416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,12288,0.04606666829850939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,6144,0.018439110782411363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,10240,0.03958400090535482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,8192,0.019425777925385367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,5120,0.0159608887301551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,7168,0.017887999614079792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,8192,0.03286844491958618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,6144,0.016230222251680162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,7168,0.029594666428036157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,5120,0.01426577733622657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,6144,0.026522666215896606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,3584,0.011159111228254108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,4096,0.012441777520709567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,256,0.00699733363257514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,3584,0.011770666473441653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,4096,0.019235556324323017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,2560,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,3584,0.017551110850440133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,3072,0.011090666883521609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,2048,0.007924444145626491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,3072,0.01569600072171953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,2560,0.010230222509966956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,1536,0.0063475556671619415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,2048,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,2048,0.012101333174440594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,1024,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,1536,0.008616888688670265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,1536,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,768,0.004528000122971005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,1024,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,1024,0.008635555704434713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,512,0.0038728887836138406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,768,0.008016889293988546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,4096,0.012808000048001608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,256,0.0031928889867332247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,512,0.007197333706749811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,3072,0.010242666635248397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,128,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,256,0.006688000013430913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,256,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,64,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,128,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,32,0.0028524444335036804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,128,0.006567111031876669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,2560,0.014075555735164218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,64,0.006298666612969504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,65536,0.1638302273220486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,32,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,51200,0.1285439994600084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,65536,0.0761786699295044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,16384,0.046689778566360474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,65536,0.1900622182422214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,768,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,51200,0.06123555368847317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,51200,0.14899555842081705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,12288,0.03311644328965081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,10240,0.027264000640975103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,5120,0.022886221607526142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,16384,0.0496782230006324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,12288,0.018985778093338013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,512,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,10240,0.017257778180970084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,7168,0.01993333299954732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,8192,0.014092443717850579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,10240,0.03336711062325372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,6144,0.01825066738658481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,8192,0.02800977892345852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,7168,0.012770666844315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,5120,0.015606222881211175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,7168,0.025451555848121643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,6144,0.01183111137813992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,4096,0.013056889176368713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,6144,0.022724444667498272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,3584,0.011807999677128263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,5120,0.019494222270117868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,4096,0.01592266725169288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,3584,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,3584,0.015227556228637695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,16384,0.023066666391160753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,3072,0.009011555876996782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,2560,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,3072,0.013464888764752282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,8192,0.02327022287580702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,2560,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,12288,0.038372443781958684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,1536,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,2560,0.012196444802814059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,2048,0.008285333712895712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,1024,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,2048,0.010713777608341642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,1536,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,768,0.00462400002612008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,1024,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,5120,0.011381333072980246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,1024,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,512,0.0038951109680864546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,768,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,4096,0.010057777994208867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,768,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,256,0.003237333355678452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,512,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,128,0.0028328889360030494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,3072,0.010424888796276517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,512,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,64,0.002601777720782492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,32,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,128,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,256,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,128,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,65536,0.14332000414530435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,64,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,2048,0.008011555506123437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,32,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,51200,0.11148267322116429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,16384,0.035826666487587824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,65536,0.18025422096252441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,51200,0.058984888924492725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,12288,0.026220444175932143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,16384,0.02204977803760105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,10240,0.022672888305452134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,16384,0.047684444321526416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,51200,0.14092800352308485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,12288,0.01829155617290073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,8192,0.020040000478426617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,12288,0.03721155391799079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,10240,0.016686222619480558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,7168,0.018065777089860704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,10240,0.032249778509140015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,8192,0.013767110804716745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,8192,0.02700088918209076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,7168,0.012553777959611682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,256,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,5120,0.013745778136783175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,7168,0.024264888630972967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,6144,0.011715555356608497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,6144,0.021500445074505273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,4096,0.011690666278203329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,5120,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,3584,0.010639111200968424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,5120,0.018235555953449674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,4096,0.009751111268997192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,65536,0.07342666387557983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,3072,0.009568889107969072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,4096,0.01574222246805827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,3584,0.009380444056457942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,2560,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,3584,0.014533332652515836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,3072,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,2048,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,3072,0.01311822235584259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,2560,0.008634666601816813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,1536,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,2048,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,2560,0.012054222325483957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,1024,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,2048,0.01035199976629681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,1536,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,1536,0.008989333278603023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,1024,0.007133333219422235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,6144,0.015895111693276297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,512,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,1024,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,256,0.0028604444944196274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,768,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,768,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,128,0.0025759999536805684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,256,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,512,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,64,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,256,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,32,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,128,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,128,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,65536,0.08769244617886013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,64,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,32,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,51200,0.08010577493243747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,65536,0.07295466793908013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,1536,0.005967111223273807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,16384,0.02717777755525377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,51200,0.058119999037848584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,12288,0.025536000728607178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,768,0.0041831110914548235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,51200,0.10971466700236003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,10240,0.02217688825395372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,16384,0.03781422310405307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,8192,0.018346667289733887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,12288,0.030157334274715845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,10240,0.016100444727473788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,512,0.006647111227114995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,7168,0.016629333297411602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,8192,0.013065777719020844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,6144,0.012883555557992725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,7168,0.012106666962305704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,7168,0.01961599952644772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,6144,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,6144,0.017628444565667045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,5120,0.010294222169452244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,4096,0.010935111178292168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,5120,0.015542222393883599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,65536,0.13839377297295466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,3584,0.010218666659461128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,16384,0.0214346663819419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,4096,0.01311911145846049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,3072,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,12288,0.017511111166742113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,3584,0.008997333546479544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,2560,0.006350222147173352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,3584,0.01237511138121287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,3072,0.008793777889675563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,10240,0.026263111167483862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,2048,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,3072,0.011190222369299995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,2560,0.008354666332403818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,8192,0.02164977788925171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,2048,0.007962666451931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,2560,0.010064000056849586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,1024,0.003882666842805015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,2048,0.009027555584907532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,768,0.0034808889031410217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,1536,0.007605333295133378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,1024,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,1536,0.00793511089351442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,512,0.0028853333658642242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,768,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,768,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,256,0.0025111111915773815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,512,0.006567111031876669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,512,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,128,0.0024746666765875286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,4096,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,256,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,64,0.0024702221983008915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,256,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,128,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,32,0.0024560000747442245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,128,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,65536,0.05977777640024821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,64,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,32,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,1536,0.004638222356637319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,51200,0.05388977792527941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,65536,0.07175199853049384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,16384,0.021943110558721755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,65536,0.1343662208980984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,51200,0.05750399827957153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,5120,0.011383111278216044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,12288,0.018244443668259513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,16384,0.020972443951500788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,10240,0.016541333662139047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,51200,0.10496711068683201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,1024,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,8192,0.014348444011476306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,12288,0.01716711123784383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,7168,0.012841777669058906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,10240,0.015831111205948722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,12288,0.028872887293497723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,8192,0.013043555948469373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,10240,0.025255999631351892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,7168,0.012100444071822695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,8192,0.020632889535692003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,5120,0.010092444717884064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,6144,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,7168,0.01899111105336083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,4096,0.008400888906584846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,6144,0.01685422162214915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,5120,0.010355555348926121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,3584,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,5120,0.014641778336630927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,4096,0.009431111315886179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,3072,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,4096,0.01276000009642707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,3584,0.0090524446633127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,2560,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,3584,0.012096000214417776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,3072,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,2048,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,3072,0.010886222124099731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,2560,0.008353778057628209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,2560,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,2048,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,1024,0.003536888708670934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,1536,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,2048,0.00871377769443724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,16384,0.03661600086424086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,768,0.003167111012670729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,1024,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,1536,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,512,0.0028106667515304354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,256,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,1024,0.006745777610275481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,768,0.00666844430896971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,6144,0.011569778124491373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,128,0.0024684444069862366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,768,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,256,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,64,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,32,0.0022124445272816554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,128,0.006224000205596288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,256,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,65536,0.04404977626270718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,64,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,32,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,65536,0.07222311364279853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,65536,0.09469244215223525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,16384,0.015365334020720588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,51200,0.05725244681040446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,768,1536,0.004577777865860197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,51200,0.07483288976881239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,12288,0.015556444724400839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,16384,0.02058844433890449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,16384,0.027067555321587458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,10240,0.014111111561457316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,12288,0.016864000095261466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,8192,0.01514755520555708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,12288,0.02158044444190131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,10240,0.01550044450494978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,7168,0.011843555503421359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,10240,0.019088000059127808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,8192,0.01279022213485506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,512,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,6144,0.010017777482668558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,8192,0.016102222932709586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,7168,0.01205599970287747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,5120,0.008947555389669206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,7168,0.01480977733929952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,6144,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,4096,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,6144,0.01349066694577535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,5120,0.010244444840484196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,51200,0.03493244449297587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,3584,0.007012444237867991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,5120,0.012058667010731168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,4096,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,3072,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,4096,0.010407110883129967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,2560,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,3072,0.00869599978129069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,3584,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,2048,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,3072,0.008986666798591614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,2560,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,2560,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,1536,0.0042168887125121225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,2048,0.00794044468137953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,1024,0.003491555651028951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,2048,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,512,0.006575999988449945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,1536,0.007407111426194509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,768,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,1536,0.006656000183688269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,1024,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,512,0.0028533333291610083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,768,0.006568000134494569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,1024,0.006265777680608962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,256,0.002469333302643564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,768,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,512,0.006321777900060018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,512,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,256,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,64,0.0021439999755885866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,128,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,128,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,3584,0.008973333570692275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,65536,0.030685332086351182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,64,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,32,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,51200,0.027356444133652583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,65536,0.093085335360633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,65536,0.0713591112030877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,16384,0.010978666444619497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,51200,0.055910223060184054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,51200,0.07380355728997125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,12288,0.011185777684052786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,128,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,16384,0.02023555503951179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,16384,0.02609511050913069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,10240,0.010327999790509542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,12288,0.01682933337158627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,256,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,8192,0.009128000173303816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,10240,0.015504888362354703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,12288,0.020991999242040846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,7168,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,10240,0.018377777602937486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,32,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,6144,0.008802666432327693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,7168,0.011904000408119626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,8192,0.015778667396969266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,5120,0.008394666843944127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,7168,0.014480888843536377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,6144,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,4096,0.007035555938879649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,3584,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,5120,0.010038221875826517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,4096,0.009303111169073317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,5120,0.011764444410800934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,3072,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,3584,0.008950221869680617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,4096,0.010336000058386061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,2560,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,3584,0.009764444496896531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,3072,0.008634666601816813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,2048,0.004594666676388847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,2560,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,1536,0.0038622220357259116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,3072,0.009134222235944536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,2560,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,2048,0.007723555796676212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,1024,0.0032266666077905228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,512,128,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,768,0.0028373334142896864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,1536,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,1024,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,512,0.002584888910253843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,768,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,1024,0.006333333336644703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,8192,0.012679999901188744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,256,0.00244355553554164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,768,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,128,0.0022053333620230355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,512,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,256,0.006180444525347815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,256,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,6144,0.013125333521101209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,64,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,256,32,0.002114666625857353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,64,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,65536,0.024304888314670984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,2048,0.007667555577225155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,65536,0.0925733314620124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,1536,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,51200,0.01775911119249132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,65536,0.0709022217326694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,16384,0.009159111314349705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,51200,0.05431110991372002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,512,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,12288,0.008243555823961893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,16384,0.019895111521085102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,128,0.006320888797442119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,16384,0.026036444637510512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,10240,0.010738666686746808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,128,0.005079999979999331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,8192,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,12288,0.016696888539526198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,12288,0.020911999874644812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,10240,0.015461333923869662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,7168,0.008718222379684448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,10240,0.018204443984561496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,8192,0.012521777715947894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,32,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,6144,0.00812888890504837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,5120,0.007344889144102733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,8192,0.015767999821239047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,7168,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,4096,0.006660444455014334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,7168,0.014440000057220459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,5120,0.010013333625263637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,3584,0.006206222292449739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,6144,0.013076444466908773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,5120,0.011704000333944956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,3072,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,4096,0.010223999619483948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,3584,0.008900444540712569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,2560,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,3584,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,3072,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,2048,0.00453955555955569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,2560,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,3072,0.008980444735950893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,1536,0.00388355553150177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,51200,0.07335199912389119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,1024,0.003152888889114062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,2048,0.007671111159854465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,2560,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,768,0.0028844444702068963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,2048,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,1536,0.007325333025720384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,512,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,1024,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,1536,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,768,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,1024,0.006119110931952794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,512,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,768,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,128,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,256,0.006261333409282897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,64,0.0018808889306253856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,128,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,256,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,32,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,6144,0.01091288857989841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,65536,0.02234488891230689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,128,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,4096,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,16384,0.009678222239017487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,32,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,12288,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,65536,0.0922551088862949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,10240,0.007618666523032718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,51200,0.07265688975652058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,8192,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,16384,0.02555822167131636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,7168,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,12288,0.020622221959961783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,6144,0.0063573333124319715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,10240,0.018224888377719455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,5120,0.008064000142945183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,8192,0.015485333071814643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,128,256,0.00240355564488305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,4096,0.00656444455186526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,3584,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,6144,0.01297155519326528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,3072,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,5120,0.011557333171367645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,2560,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,4096,0.010069333016872406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,2048,0.0041946665280395085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,3584,0.009639110830095079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,51200,0.01726222203837501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,1536,0.003680888977315691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,3072,0.009022222624884712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,1024,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,64,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,768,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,2560,0.008069333102968004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,512,0.00254133323000537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,2048,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,256,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,1536,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,128,0.0019102222803566193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,1024,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,64,0.0017973333597183228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,768,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,64,32,0.001826666709449556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,512,0.005336888962321811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,65536,0.01888533267709944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,256,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,128,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,64,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,16384,0.01518311103185018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,32,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,7168,0.014170666535695394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,12288,0.013091555900043912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,10240,0.007818666597207388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,65536,0.09222400188446045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,512,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,51200,0.07277955611546834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,8192,0.010178666975763109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,7168,0.017909333109855652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,16384,0.025654221574465435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,6144,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,12288,0.020658666888872784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,5120,0.013558222187889947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,4096,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,8192,0.015504888362354703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,7168,0.014170666535695394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,3584,0.006785777707894643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,3072,0.006196444647179709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,6144,0.013079110946920184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,2560,0.005499555418888728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,4096,0.010056888891590966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,2048,0.004579555657174853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,3584,0.009567110902733272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,1536,0.003997333347797394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,3072,0.008831111093362173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,1024,0.0031911111954185697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,2560,0.00814222213294771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,768,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,2048,0.0075911109646161394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,51200,0.012730666332774691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,512,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,256,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,1536,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,128,0.0018595556418100994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,1024,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,64,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,768,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,769,32,32,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,256,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,128,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,64,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,10240,0.018039999736679923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,32,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,5120,0.011689778003427716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,16384,0.8752915594312879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,512,0.0052879999081293745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,16384,1.0291129218207467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,12288,0.6483022371927897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,16384,0.3254888852437337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,51200,1.09770573510064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,12288,0.7416106859842936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,12288,0.2519520123799642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,51200,2.9980470869276257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,16384,0.8901537789238824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,10240,0.7855742242601184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,10240,0.6133688820732964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,8192,0.5055040253533257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,51200,3.489765379163954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,10240,0.21238311131795248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,7168,0.415359099706014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,12288,0.6574666765001085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,8192,0.17016532686021593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,8192,0.7065084245469836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,10240,0.543815983666314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,6144,0.3641840087042914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,7168,0.15392088890075684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,8192,0.4360364543067084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,5120,0.30253956052992076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,6144,0.3688364558749729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,6144,0.13313066959381104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,4096,0.26628798908657497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,5120,0.11639732784695095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,7168,0.382259554333157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,5120,0.3140960004594591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,6144,0.3338942262861464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,4096,0.0955315563413832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,3584,0.19581600030263266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,7168,0.42778399255540633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,4096,0.3133768770429823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,5120,0.28597956233554417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,3072,0.22057955794864229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,3584,0.08438311020533244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,4096,0.23173067304823133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,3584,0.215521772702535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,3072,0.07630044221878052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,2560,0.14915288819207087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,3072,0.18962932957543266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,3584,0.20224355326758492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,2048,0.1415448851055569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,2560,0.15982488791147867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,3072,0.18884266747368705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,1536,0.10826310846540664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,2048,0.13198933336469862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,2048,0.05697066916359795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,2560,0.1513813336690267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,1024,0.0738453335232205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,1536,0.04520710971620348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,2048,0.1271457804573907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,1536,0.10936711231867473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,768,0.05318311187956068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,1536,0.10050311353471543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,1024,0.03734577695528666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,1024,0.07965866724650066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,512,0.039623111486434937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,768,0.03299999899334378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,768,0.0649413333998786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,1024,0.0787315567334493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,256,0.02828888760672675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,512,0.05590044127570259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,2560,0.06686133146286011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,768,0.0660159985224406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,128,0.022633777724372015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,256,0.050776888926823936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,256,0.02773511078622606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,51200,2.96577517191569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,128,0.04748622245258755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,64,0.02438755498992072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,128,0.027639999985694885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,256,0.0429440008269416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,32,0.024616888827747766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,64,0.04684533344374763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,32,0.04680266645219591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,512,0.03014577759636773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,512,0.054607109891043774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,128,0.03407111101680332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,65536,1.008696026272244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,51200,2.541224797566732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,65536,3.2096159193250866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,51200,0.8198391066657172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,16384,1.0078168445163302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,51200,2.6885155571831594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,65536,3.491642634073893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,16384,0.3100373215145535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,16384,0.7831670973036023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,12288,0.6844133271111382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,12288,0.5675466855367025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,12288,0.19950932926601836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,10240,0.43999467955695254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,10240,0.1701271136601766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,10240,0.49856890572441953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,12288,0.5123244391547309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,16384,0.6934417618645562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,8192,0.38279022110833066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,7168,0.3513306776682536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,10240,0.4397884474860297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,8192,0.13666755623287624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,8192,0.34842488500807023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,7168,0.3333031071556939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,8192,0.363334231906467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,6144,0.29273777537875706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,5120,0.24299555354648164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,6144,0.10865600241555108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,6144,0.2985688845316569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,5120,0.09145066473219131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,6144,0.2685573365953234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,7168,0.310023996565077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,5120,0.2651040024227566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,4096,0.22174755732218424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,51200,2.364651574028863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,7168,0.12423022588094075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,5120,0.22668798764546713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,3584,0.1902986632453071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,4096,0.07554488711886935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,4096,0.19502666261461046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,3072,0.1574880017174615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,4096,0.1859217749701606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,3584,0.06897066699133979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,2560,0.1390088929070367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,3072,0.14821778403388128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,3584,0.1777813302146064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,3584,0.1657760010825263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,2048,0.11084622144699097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,2560,0.1270186636182997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,2560,0.05298044284184774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,1536,0.08233155806859334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,65536,3.0258053673638234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,2048,0.04555911156866285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,2048,0.10456533564461602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,2560,0.1215546660953098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,1024,0.05145600106981066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,1536,0.08273155821694268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,2048,0.10087200005849202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,1024,0.06379199690288968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,1024,0.03015822172164917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,768,0.042641778786977134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,1536,0.09292622407277425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,1024,0.06495910882949829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,768,0.029708445072174072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,3072,0.07134399811426799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,512,0.0317520002524058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,768,0.052129778597089976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,256,0.02195555633968777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,768,0.054029332266913525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,512,0.02516355613867442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,512,0.04499377806981405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,128,0.018224888377719455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,256,0.02323644525474972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,256,0.04087200098567539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,512,0.046458668178982206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,64,0.01610488858487871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,128,0.022934221559100684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,256,0.03641866644223531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,1536,0.03633155425389608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,64,0.03769866625467936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,128,0.02836622132195367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,32,0.03767022159364488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,3072,0.14356177382998997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,65536,0.8973307079739041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,65536,0.3673866589864095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,128,0.038316445218192205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,32,0.01682577696111467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,65536,1.079067548116048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,51200,0.7112906773885092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,16384,0.23642312155829537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,16384,0.09228355354732937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,16384,0.25832266277737087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,51200,0.2861057652367486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,51200,0.8343244658576118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,12288,0.21244800090789795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,10240,0.1563697788450453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,12288,0.07125866413116455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,12288,0.18882755438486734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,16384,0.24718043539259169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,10240,0.15670577685038248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,8192,0.12619377507103816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,12288,0.1842391093571981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,10240,0.060601777500576444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,7168,0.13434488243526882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,8192,0.1307013299730089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,65536,1.0325626797146268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,10240,0.15552977720896402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,8192,0.1271795564227634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,7168,0.04503644506136576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,7168,0.11317600144280328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,6144,0.11011111074023777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,5120,0.09832711352242364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,51200,0.8059795697530111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,6144,0.03973599937227037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,6144,0.09818133380677964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,4096,0.06573333342870076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,5120,0.09161333243052165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,5120,0.040194667047924466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,6144,0.09903733597861396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,3584,0.06629155741797553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,4096,0.02899555696381463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,5120,0.08620444271299575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,4096,0.06827733251783583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,3072,0.056332442495557994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,3584,0.026557332939571802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,3584,0.060324443711174854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,4096,0.0734951098759969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,2560,0.04743822084532844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,8192,0.0495377779006958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,3584,0.06651466422610812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,3072,0.05313422282536825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,2048,0.03826133410135905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,2560,0.02160266704029507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,2560,0.046126223272747464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,3072,0.05836977561314901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,1536,0.030271109607484605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,2048,0.018808888064490426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,2048,0.03839288817511664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,2560,0.05093422200944689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,7168,0.1277137729856703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,1536,0.016176889340082806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,1536,0.030761778354644775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,2048,0.04391111267937554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,768,0.016516443755891588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,1024,0.01367911116944419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,1536,0.037512888511021934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,512,0.01275288893116845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,768,0.012445333103338877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,768,0.02030044462945726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,768,0.026351110802756414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,512,0.017237333787812125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,3072,0.024180443750487432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,512,0.022389334109094407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,256,0.01519377695189582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,128,0.009398221969604492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,256,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,1024,0.020614221692085266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,256,0.017854221993022494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,128,0.014674666855070325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,128,0.0107360002067354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,64,0.00869777798652649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,32,0.008938666847017076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,1024,0.024068444967269897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,128,0.015191111299726697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,1024,0.030363556411531236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,64,0.014765333798196582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,32,0.01462044484085507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,512,0.01129155523247189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,256,0.011353777514563667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,65536,0.6849608951144748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,51200,0.5017537540859646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,65536,0.27078933186001247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,16384,0.15808978345659044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,65536,0.8126062287224664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,51200,0.6424826516045464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,51200,0.20546221733093262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,16384,0.06816089153289795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,12288,0.14760355154673258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,16384,0.17755911085340711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,10240,0.10266755686865912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,12288,0.05292888813548618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,12288,0.13396533330281576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,51200,0.5429351064893935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,8192,0.09752088785171509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,65536,0.6905324194166395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,10240,0.045202665858798556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,10240,0.12243555651770698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,7168,0.07366044653786553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,8192,0.03706311186154684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,10240,0.11367644204033746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,8192,0.10832266675101386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,16384,0.20508978101942274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,6144,0.07250133487913343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,7168,0.03324355681737264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,8192,0.09767555528216892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,7168,0.08819288677639431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,5120,0.06257510847515531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,6144,0.029311110575993855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,7168,0.08597066667344834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,4096,0.053710222244262695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,6144,0.07568977938758002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,5120,0.06544799937142266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,5120,0.02549244463443756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,3584,0.03783822059631348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,4096,0.021900445222854614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,4096,0.05433155430687798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,3072,0.03874399926927354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,5120,0.06520177920659383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,3584,0.020008888509538438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,4096,0.054851555162005954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,2560,0.03219288918707106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,3072,0.01830577850341797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,3584,0.050184001525243126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,3072,0.042345778809653394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,12288,0.1308257712258233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,2048,0.02583911187118954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,3072,0.044807109567854136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,2560,0.03631911012861464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,1536,0.019863999552196927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,2048,0.030586666531032983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,2048,0.015263110399246216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,6144,0.07612266805436876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,1024,0.014463111758232117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,1536,0.013395555317401886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,2048,0.03472800056139628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,1536,0.024822221861945257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,768,0.011391999820868174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,1024,0.011404444773991903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,1024,0.019766221443812054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,1536,0.028824888997607764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,512,0.008954666554927826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,768,0.016571554872724745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,3584,0.047998223039839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,768,0.010198222266303169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,1024,0.024029332730505202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,256,0.0064497776329517365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,512,0.009684444301658207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,512,0.014044443766276041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,128,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,256,0.008954666554927826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,512,0.018792000081804063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,2560,0.016892444756295946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,64,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,256,0.015976000163290236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,32,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,128,0.009113777842786577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,2560,0.03963288995954726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,64,0.012138666378127204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,128,0.012858666479587555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,32,0.012095111111799875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,65536,0.6145244704352485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,65536,0.24718933635287813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,51200,0.44475200441148544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,768,0.02123111155298021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,65536,0.6828186776902941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,256,0.012744888663291931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,16384,0.15275822745429143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,128,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,51200,0.19129244486490884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,16384,0.06237688991758558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,16384,0.16438666979471842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,51200,0.5489813486735026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,12288,0.12798133161332872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,16384,0.1689822276433309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,10240,0.09719110859764947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,12288,0.04989777670966255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,12288,0.12342133786943223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,8192,0.08296444680955675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,10240,0.04262400004598829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,10240,0.10411289003160264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,12288,0.1298604408899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,65536,0.6792106628417969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,7168,0.06450488832261828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,51200,0.5214906798468696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,8192,0.08457689152823554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,8192,0.03476000163290236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,10240,0.11096711291207208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,6144,0.06298310889138116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,7168,0.07795111338297527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,5120,0.048023998737335205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,6144,0.027533332506815594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,6144,0.0658444431093004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,7168,0.08211644490559895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,4096,0.04008177916208903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,5120,0.024471999870406255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,6144,0.07182488838831584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,5120,0.05582222011354235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,3584,0.03954844342337714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,4096,0.021149333980348375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,5120,0.06343999836179945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,4096,0.045964443021350436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,3072,0.034671111239327326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,3584,0.04079733292261759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,3584,0.019350222415394254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,4096,0.05361066593064202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,2560,0.03056000007523431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,3072,0.017888888716697693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,3584,0.047404444879955716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,3072,0.03584622343381246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,7168,0.03196355700492859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,2048,0.025128000312381323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,2560,0.03162044286727905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,8192,0.09162222014533149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,2560,0.016559999850061204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,3072,0.04446844591034783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,2048,0.014814222852389017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,1024,0.015191111299726697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,2560,0.03834755553139581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,2048,0.02665600015057458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,1536,0.012680889003806643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,768,0.011897777517636618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,1536,0.021929777330822412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,1024,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,1024,0.017557332913080852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,512,0.009024889104896123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,1536,0.028008000718222723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,768,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,1024,0.02253599961598714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,256,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,512,0.012448888685968188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,768,0.020202666521072388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,512,0.01740711099571652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,128,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,256,0.008964444200197855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,1536,0.02001422146956126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,256,0.0147724449634552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,64,0.005711110929648082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,128,0.011026666396194033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,32,0.005807999935415056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,128,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,128,0.012108444339699216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,2048,0.03300088975164626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,64,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,32,0.011009777585665384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,768,0.014551111393504672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,65536,0.47796090443929035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,512,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,65536,0.19351911544799805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,256,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,16384,0.13439733452267116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,65536,0.5361893441942003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,51200,0.15611288282606336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,51200,0.43812889522976345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,16384,0.051071110698911876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,12288,0.10822400119569565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,16384,0.13482222292158338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,10240,0.08229421907001071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,12288,0.04089422358406915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,16384,0.14072799682617188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,12288,0.10249688890245225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,8192,0.07424889008204143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,65536,0.5186986923217773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,10240,0.03523911039034525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,51200,0.40344622400071883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,10240,0.08701066838370429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,12288,0.10414666599697536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,7168,0.06693600283728705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,8192,0.028809779220157202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,8192,0.07112533516354032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,10240,0.08987555238935684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,6144,0.05544622076882256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,7168,0.02886755598915948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,7168,0.06387822495566474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,8192,0.07689422369003296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,5120,0.04474222328927782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,6144,0.026151110728581745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,6144,0.05500088797675239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,7168,0.06666222545835707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,4096,0.03650577863057455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,5120,0.0206675562593672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,5120,0.048298666874567665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,6144,0.05856000052558052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,3584,0.032051556640201144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,4096,0.018254222141371835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,5120,0.05259199937184652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,4096,0.039056890540652804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,3072,0.027377777629428442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,3584,0.016584888100624084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,3584,0.03496444556448195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,4096,0.04413511024581062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,2560,0.02351644469632043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,3072,0.015188443991872998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,3072,0.03127022253142463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,3584,0.04026044408480326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,2048,0.019753777318530612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,51200,0.34305689069959855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,3072,0.03626222080654568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,2560,0.01424800025092231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,2560,0.027414222558339436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,1536,0.014755555325084262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,2560,0.032010667853885226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,2048,0.023357333408461675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,2048,0.01275288893116845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,1536,0.019094222121768527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,2048,0.028120001157124836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,768,0.010264889233642155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,1024,0.015126221709781222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,1536,0.024375999967257183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,1024,0.009668444593747457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,512,0.0083137775460879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,1024,0.020054222808943856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,768,0.0127688886390792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,512,0.011152889165613385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,256,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,768,0.01773422294192844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,512,0.008601778083377415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,512,0.01536622146765391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,256,0.008047999607192146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,128,0.005696889013051987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,256,0.010140444669458602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,256,0.01426577733622657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,1024,0.012072888513406118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,64,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,128,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,128,0.011149333582984077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,64,0.009727111293209923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,32,0.009747555686367882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,768,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,65536,0.4443813429938422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,65536,0.18229511049058703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,51200,0.3583653238084581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,65536,0.4920319981045193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,16384,0.1123466624153985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,51200,0.38907289505004883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,128,0.008029333419269985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,1536,0.011008888483047485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,51200,0.1448026630613539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,32,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,12288,0.0891431106461419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,16384,0.048433777358796865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,16384,0.12078844176398383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,65536,0.5155786408318414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,12288,0.038444442881478205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,12288,0.09433688720067342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,16384,0.13259466489156088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,8192,0.07960533433490329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,51200,0.3956257767147488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,10240,0.033207999335394964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,10240,0.07859110832214355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,12288,0.1033431092898051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,7168,0.06931733422809176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,8192,0.02865955564710829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,8192,0.06480711036258273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,6144,0.05995911359786987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,7168,0.024580443898836773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,7168,0.05749333567089505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,8192,0.0732426643371582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,5120,0.04620711008707682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,6144,0.022036444809701707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,7168,0.06601422362857394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,6144,0.05010933346218533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,4096,0.03843555516666836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,5120,0.01994311147265964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,5120,0.042986667818493314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,3584,0.033719112475713096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,4096,0.01759377784199185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,10240,0.09709244304233128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,4096,0.03584355447027419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,5120,0.05181688732571072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,3072,0.028029332558314007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,4096,0.04337866769896614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,3584,0.016167999969588388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,3584,0.03242311212751601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,2560,0.02460177739461263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,3072,0.015035554766654968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,3584,0.039593779378467135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,3072,0.028778665595584448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,2048,0.01960444450378418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,10240,0.08830755286746556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,2560,0.025416889124446448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,2560,0.013802666631009845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,3072,0.03513155712021722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,1536,0.01665155589580536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,2048,0.012327999704413943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,2560,0.031480888525644936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,1024,0.010071111222108206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,1536,0.017678222722477384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,2048,0.02773511078622606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,6144,0.05916355715857612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,768,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,1536,0.023732443650563557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,1024,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,1024,0.0138586668504609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,512,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,768,0.00885511106914944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,768,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,512,0.01036800030204985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,256,0.004998222407367495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,768,0.017553778158293832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,512,0.008365333080291748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,256,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,128,0.00479911102188958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,256,0.009584888815879822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,2048,0.02155911094612545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,256,0.014150222142537435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,64,0.0042053332759274375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,128,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,1536,0.010440888504187265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,32,0.004237333519591225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,128,0.010971555279360877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,64,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,32,0.009308444129096137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,1024,0.019326221611764696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,65536,0.1447111103269789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,51200,0.26124800576104057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,65536,0.42374576462639707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,512,0.015683554940753512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,16384,0.10763911406199138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,51200,0.33715467982821995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,51200,0.11265244748857285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,16384,0.10515822304619683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,12288,0.07879466480678983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,128,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,16384,0.03832355472776625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,65536,0.37685510847303605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,10240,0.05572355455822415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,12288,0.030064887470669214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,16384,0.10125955608155991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,12288,0.08012888828913371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,51200,0.3018595642513699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,65536,0.3715084393819173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,10240,0.067994667424096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,10240,0.02646488944689433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,12288,0.07898666461308797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,7168,0.04381688767009311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,8192,0.021525333325068157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,8192,0.0554346674018436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,6144,0.038881778717041016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,10240,0.06856711043251885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,7168,0.019910222954220243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,5120,0.032519999477598406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,8192,0.05778933233684964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,6144,0.018163555198245578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,6144,0.0432168874475691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,7168,0.05192710955937704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,4096,0.026333333717452154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,5120,0.0162426663769616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,6144,0.046485334634780884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,3584,0.02273688879277971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,4096,0.01460800071557363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,4096,0.0314026673634847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,5120,0.04054222173160977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,3072,0.020185778538386028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,3584,0.013843555417325763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,8192,0.05068177647060818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,3584,0.028323554330401953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,2560,0.016869333055284288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,3072,0.025137777129809063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,3072,0.01240711079703437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,3584,0.03164266546567281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,2048,0.014130666851997375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,2560,0.011415111521879831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,3072,0.028582221931881372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,7168,0.04973155591222975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,1536,0.011789333489206104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,2048,0.010018666585286459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,2560,0.025659556190172832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,5120,0.03756622142261929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,1536,0.015493333339691162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,1536,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,2048,0.02296977738539378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,1024,0.008972444468074376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,1024,0.008361777497662438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,1536,0.020391111572583515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,768,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,1024,0.012404444317022959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,4096,0.03461600012249417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,512,0.006072000082996156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,1024,0.01722044414944119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,768,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,768,0.011226666470368704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,256,0.004860444615284602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,512,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,512,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,768,0.015192000402344598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,128,0.00397866674595409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,2560,0.022239999638663396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,256,0.007359111474619971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,256,0.008963555925422246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,512,0.013774221969975365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,64,0.003559999995761447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,2048,0.018964444597562153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,128,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,128,0.008656889200210571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,256,0.012390221986505719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,128,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,64,0.008674666285514832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,32,0.008643555144468943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,65536,0.347941345638699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,51200,0.2607288890414768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,65536,0.13225066661834717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,65536,0.36716887685987687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,16384,0.09034133619732326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,51200,0.28329955206977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,51200,0.10471733411153157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,12288,0.06724533107545641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,16384,0.09212177991867065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,65536,0.36725600560506183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,51200,0.29245421621534556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,16384,0.10193422105577256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,12288,0.029345777299669053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,10240,0.05639288822809855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,12288,0.07372177971733941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,8192,0.04579910967085096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,10240,0.02548533346917894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,12288,0.07867199844784208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,10240,0.060637334982554116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,7168,0.040017777019076876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,8192,0.04907288816240099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,8192,0.02125777800877889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,10240,0.06744444370269775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,6144,0.03503111004829407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,7168,0.01942044496536255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,7168,0.04452177882194519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,8192,0.05538488758934868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,5120,0.029072887367672388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,6144,0.017870222528775532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,6144,0.03832622369130453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,4096,0.02381155557102627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,7168,0.0503057771258884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,16384,0.03649866580963135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,5120,0.016122667325867545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,5120,0.03307644526163737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,6144,0.04561777909596761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,3584,0.021346666746669348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,4096,0.014175999495718213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,3072,0.01891288823551602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,5120,0.03983822133806016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,4096,0.028130667077170476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,3584,0.013091555900043912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,2560,0.01649777756796943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,3584,0.025591111845440332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,4096,0.03390044305059645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,3072,0.012051555845472546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,2048,0.014092443717850579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,3072,0.023071999351183575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,3584,0.03114933437771267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,2560,0.011409777734014722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,2560,0.02032177812523312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,1536,0.011762667033407422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,3072,0.02847555610868666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,2048,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,1024,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,2048,0.017284444636768766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,1536,0.009027555584907532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,1536,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,2048,0.0224906669722663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,768,0.007140444384680853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,1024,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,1024,0.011407111254003314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,1536,0.0196435550848643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,512,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,768,0.008006222546100616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,1024,0.016153777639071148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,256,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,32,0.003847111016511917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,512,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,512,0.007637333538797166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,512,0.013722666435771518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,256,0.007325333025720384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,256,0.008301333420806462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,128,0.004239999999602635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,256,0.012262221839692859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,128,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,64,0.00388355553150177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,128,0.007707555260923173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,32,0.00453599997692638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,2560,0.02499466637770335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,128,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,64,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,65536,0.25116001235114205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,32,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,51200,0.1903404394785563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,65536,0.12441333134969075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,768,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,65536,0.31347378094991046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,16384,0.062295112344953746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,768,0.01480888823668162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,51200,0.24179556634691027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,51200,0.09868355592091878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,12288,0.056622220410241016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,16384,0.0795582201745775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,65536,0.3608284526401096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,10240,0.047615110874176025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,12288,0.028005331754684448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,12288,0.06133955717086792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,16384,0.0987608896361457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,51200,0.2840924527910021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,8192,0.034741332133611046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,10240,0.024855110380384658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,12288,0.07703999678293864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,7168,0.03334844443533156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,10240,0.05254488852288988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,8192,0.020647111866209242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,8192,0.04323377874162462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,10240,0.06551466385523479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,6144,0.029081778393851385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,7168,0.018921777606010437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,7168,0.03847555650605096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,8192,0.0550177792708079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,5120,0.024456888437271118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,6144,0.01733688844574822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,7168,0.0498888889948527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,4096,0.019984000258975558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,5120,0.015458666616015963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,6144,0.044255110952589244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,16384,0.03535200158754984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,3584,0.018000000052981906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,5120,0.03886399997605218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,4096,0.025257777836587694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,3072,0.015814221567577787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,3584,0.012867555850081973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,3584,0.022996443841192458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,4096,0.03344088792800903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,2560,0.013810666898886362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,3584,0.030753778086768255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,3072,0.020383111304706998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,2560,0.010463111102581024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,2560,0.01824355622132619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,2048,0.012356444365448423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,3072,0.027656000521447923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,2560,0.02472266720400916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,6144,0.034287999073664345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,1536,0.009011555876996782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,2048,0.015398222539159985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,5120,0.02955377764172024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,1024,0.007230222225189209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,2048,0.02208711041344537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,1536,0.008953777452309927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,1536,0.012625777886973487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,768,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,4096,0.013805333111021252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,1024,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,1536,0.018573333819707234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,1024,0.01039644413524204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,512,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,768,0.007890666524569193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,1024,0.015850666496488783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,256,0.004181333300140169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,512,0.008114666574531132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,512,0.007385777930418651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,768,0.01479111115137736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,128,0.0035048888789282907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,512,0.013412444127930535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,256,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,2048,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,128,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,128,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,256,0.012039999994966718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,128,0.010025777750545079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,64,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,32,0.007256000406212277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,65536,0.20014845000372994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,768,0.009321778184837764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,3072,0.011457777685589261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,65536,0.2781253390842014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,65536,0.12246578269534641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,51200,0.20549866888258192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,256,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,16384,0.06883466906017728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,51200,0.09820266564687093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,51200,0.2157448927561442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,64,0.0034702221552530923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,32,0.003540444291300244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,12288,0.05004622207747566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,16384,0.034435556994544134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,16384,0.07044088840484619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,10240,0.04379466507169935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,51200,0.2821733421749539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,12288,0.027588443623648748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,12288,0.07645955350663927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,10240,0.04612000121010674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,8192,0.03340355555216471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,10240,0.06534400251176622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,8192,0.03821422325240241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,8192,0.02016177773475647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,7168,0.030188444587919448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,8192,0.05485244592030843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,65536,0.35857155587938094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,6144,0.02511022157139248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,7168,0.03458844290839301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,7168,0.018602665927675035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,5120,0.02186222208870782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,7168,0.04901688959863451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,6144,0.016872000363137987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,6144,0.0303422212600708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,12288,0.05446044603983561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,4096,0.019481778144836426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,5120,0.015079999963442484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,5120,0.02623644471168518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,3584,0.014888889259762235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,10240,0.02437244521247016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,6144,0.044138666656282216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,4096,0.013341333303186627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,4096,0.022648889157507155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,5120,0.03860444492763943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,3584,0.012320000264379712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,3584,0.020385776956876118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,2560,0.01163111130396525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,3072,0.01131199962562985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,3072,0.018440888987647164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,3584,0.030513776673210993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,2048,0.010068444742096795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,2560,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,3072,0.027043556173642475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,2048,0.013880000346236758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,2048,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,2560,0.024707555770874023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,1536,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,16384,0.09869244363572861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,1536,0.008725333544943068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,1024,0.006341333190600078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,1536,0.011418666276666852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,1536,0.018538667096032035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,1024,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,768,0.0052933332820733385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,3072,0.01367022262679206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,4096,0.032736000087526106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,1024,0.015830222103330824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,512,0.004509333521127701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,768,0.007725333174069722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,768,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,256,0.003879111260175705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,512,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,768,0.014434667097197639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,512,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,2560,0.015856888559129503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,128,0.003154666680428717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,256,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,256,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,512,0.013183111117945777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,64,0.003071111109521654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,32,0.0031902222997612427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,128,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,128,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,256,0.012475555141766867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,2048,0.021314667330847845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,128,0.010000000397364298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,32,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,65536,0.22311645083957246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,51200,0.13312799400753444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,1024,0.009688888986905416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,65536,0.2386577659183078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,16384,0.05560266971588135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,51200,0.06639733579423693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,51200,0.18609600596957734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,65536,0.22447734408908418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,12288,0.03942044576009115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,16384,0.06100710895326403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,16384,0.023980443676312763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,51200,0.1945635610156589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,10240,0.031006223625606958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,12288,0.019253333409627277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,16384,0.06517599688635932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,12288,0.046590222252739795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,8192,0.028060442871517603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,10240,0.017231111725171406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,10240,0.04047822289996677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,12288,0.051496889856126576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,7168,0.02422044508987003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,8192,0.03319022059440613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,8192,0.014464888307783337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,64,0.0070159998204973005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,10240,0.044920888211992055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,6144,0.02168088820245531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,8192,0.03836177786191305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,7168,0.013444444371594323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,5120,0.01869244376818339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,7168,0.030249777767393324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,6144,0.026935110489527386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,6144,0.012252444194422828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,4096,0.015503111812803479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,7168,0.03461866577466329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,6144,0.030974222554100886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,5120,0.02325422234005398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,3584,0.013972444666756524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,5120,0.028033776415718928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,4096,0.020066666934225295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,3584,0.01791200041770935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,4096,0.024301333559883967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,3072,0.012392000191741519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,3584,0.021936888496081035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,3072,0.016393777396943834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,2560,0.011036444041464063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,3072,0.009081777599122789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,3072,0.019890666007995605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,65536,0.08261155419879489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,2048,0.009415110780133141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,2560,0.014130666851997375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,2560,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,1536,0.007109333243634965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,5120,0.011044444309340583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,2048,0.008371555142932469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,2048,0.01241422196229299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,1024,0.005485333502292633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,2560,0.018248000078731112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,4096,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,1536,0.007672888537247975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,2048,0.016195555528004963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,768,0.004552000098758274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,1024,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,3584,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,1024,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,512,0.004233777936961916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,1536,0.01481066644191742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,768,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,768,0.0070053330726093715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,1024,0.013157332936922709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,256,0.003330666571855545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,512,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,512,0.007258666886223688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,128,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,768,0.012465777496496836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,256,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,64,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,256,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,512,0.011511999699804517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,128,0.0063928887248039246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,128,0.006649777707126405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,256,0.010673777924643623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,64,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,65536,0.18890666961669922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,128,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,32,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,51200,0.1330880059136285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,65536,0.07881066534254286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,1536,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,16384,0.05041244294908312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,65536,0.22513156467013887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,51200,0.0610062215063307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,51200,0.17580710517035592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,12288,0.036580443382263184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,16384,0.02311822275320689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,65536,0.2195075617896186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,16384,0.05849066707823011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,51200,0.1731004450056288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,12288,0.019140443868107267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,16384,0.06286844280030993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,8192,0.024646222591400146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,10240,0.01726222203837501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,10240,0.03901422354910109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,12288,0.050276445017920605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,7168,0.021398221453030903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,8192,0.014152889450391134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,8192,0.03247822324434916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,7168,0.029491557015313044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,7168,0.01311555587583118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,8192,0.0381297767162323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,6144,0.018730666902330186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,6144,0.012112889024946423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,7168,0.033722665574815534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,5120,0.016167999969588388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,6144,0.026195555925369263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,10240,0.030810667408837214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,5120,0.011011555790901184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,4096,0.013762666947311826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,5120,0.022470222579108343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,12288,0.04540888799561394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,5120,0.027580445011456806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,3584,0.012479110724396177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,32,0.0028524444335036804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,10240,0.04428799947102865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,3584,0.01702933344576094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,3072,0.011261333194043903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,3584,0.009394666386975182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,3584,0.021470222208235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,3072,0.01554844445652432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,3072,0.008943999807039896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,2560,0.009946666657924652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,3072,0.01962933275434706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,6144,0.030741333961486816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,2048,0.008662222160233391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,2560,0.013800000150998434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,2560,0.008640888664457533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,4096,0.009928888744778102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,2048,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,1536,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,2048,0.012087110843923358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,4096,0.023876445160971746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,2048,0.016016888949606154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,1536,0.007626666790909237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,1024,0.005462222215202119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,1536,0.010173333187898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,768,0.004466666529575984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,1024,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,1536,0.01427288850148519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,1024,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,512,0.0038453332251972626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,768,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,768,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,1024,0.01276088919904497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,256,0.0034328889515664843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,512,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,512,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,768,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,128,0.002891555635465516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,256,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,512,0.011443555355072021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,2560,0.01776711146036784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,4096,0.018927999668651156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,64,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,128,0.006607111129495833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,32,0.0028355556229750314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,128,0.006308444258239534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,256,0.010737777584128909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,65536,0.13411911328633627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,128,0.008655110994974772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,64,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,32,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,65536,0.07303910785251193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,51200,0.10655999845928615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,16384,0.04253510965241326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,65536,0.18420622083875868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,51200,0.05912444326612684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,12288,0.03104355600145128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,16384,0.02203111184967889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,51200,0.14476444986131456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,65536,0.21693600548638237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,16384,0.04888977938228183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,10240,0.025803556044896443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,12288,0.01882222294807434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,51200,0.17075110806359184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,16384,0.06275466415617201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,8192,0.02050044470363193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,12288,0.03825866513782077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,256,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,10240,0.0329751107427809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,10240,0.016584888100624084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,7168,0.018381334013409086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,12288,0.049999998675452344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,6144,0.015956444872750174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,8192,0.02781244450145298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,10240,0.04309066798951891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,7168,0.025172443853484258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,5120,0.013738666971524557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,8192,0.036469333701663546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,6144,0.011425777441925473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,6144,0.022379555635982092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,4096,0.011556444068749746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,5120,0.010399111443095738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,5120,0.019202666150199044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,4096,0.015850666496488783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,5120,0.02608266638384925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,4096,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,3584,0.010653333531485664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,3584,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,4096,0.022737777895397607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,3072,0.0096142225795322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,3584,0.01481599940194024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,3072,0.008736888567606607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,3072,0.01347733371787601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,7168,0.012706666356987424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,3072,0.01907822158601549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,2560,0.012108444339699216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,7168,0.033366223176320396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,2048,0.00742133292886946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,2560,0.008683555656009251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,6144,0.029959999852710303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,2560,0.017510222064124215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,2048,0.008214222060309516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,1536,0.006506666541099548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,2048,0.010616888602574667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,1536,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,2048,0.015842666228612263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,1024,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,1536,0.009015999734401703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,1536,0.014151111245155334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,1024,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,768,0.004576888763242298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,1024,0.007890666524569193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,3584,0.021182222498787776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,512,0.00388711111413108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,1024,0.012709333664841123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,768,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,768,0.007143110864692264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,256,0.0032337777730491427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,512,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,512,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,768,0.011724444727102915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,128,0.0031999999450312722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,256,0.006261333409282897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,256,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,512,0.011417778001891242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,64,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,128,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,128,0.005979555762476391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,256,0.010061333576838175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,32,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,65536,0.12335999806722005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,64,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,128,0.008400888906584846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,32,0.006067555397748947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,51200,0.08684977557924058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,65536,0.14293955432044134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,65536,0.07293688588672213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,8192,0.013755555782053204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,2560,0.008626666333940294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,16384,0.03230311142073737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,51200,0.05927199787563748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,51200,0.11365155378977458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,16384,0.021496888664033677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,12288,0.02288355595535702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,16384,0.03874399926927354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,10240,0.019367110398080613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,12288,0.01807733376820882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,51200,0.16227911578284368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,16384,0.05901955233679878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,12288,0.03073422114054362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,8192,0.01534577707449595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,10240,0.015955555770132277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,10240,0.026603556341595117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,12288,0.04664177695910136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,7168,0.014035556051466199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,8192,0.013052444491121503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,8192,0.022477333744366963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,6144,0.012423111332787408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,10240,0.04067911042107476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,7168,0.01204622205760744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,5120,0.010673777924643623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,7168,0.020434666011068556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,8192,0.03475111060672336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,6144,0.011146666275130378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,6144,0.018546667363908555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,7168,0.03195733494228787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,5120,0.01035377797153261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,5120,0.015633778439627755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,3584,0.008303110798199972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,6144,0.02871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,4096,0.009647111097971598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,4096,0.01312355531586541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,5120,0.02498133314980401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,3072,0.007348444726732042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,65536,0.20550310611724854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,3584,0.009327111144860586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,2560,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,3584,0.012461333639091916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,3072,0.008730666504965888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,3072,0.01164533363448249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,3584,0.020204444726308186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,2048,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,2560,0.008319999608728621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,2560,0.010336000058386061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,3072,0.018593778212865193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,1536,0.004862222406599256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,2048,0.008017777568764156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,2560,0.016920000314712524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,2048,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,1024,0.0038648889296584656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,1536,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,1536,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,2048,0.015771556231710646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,768,0.003540444291300244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,4096,0.00927022182279163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,1536,0.013807111316257052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,1024,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,512,0.0031555555760860443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,768,0.006983111302057902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,768,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,1024,0.012720000412729053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,256,0.0026435556097163092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,512,0.006492444624503453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,768,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,4096,0.021894221504529316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,128,0.0024968888610601425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,512,0.011379555695586734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,256,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,256,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,64,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,128,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,32,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,256,0.010420444111029306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,128,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,64,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,65536,0.09085066450966729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,128,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,32,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,51200,0.07223555776807997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,65536,0.07308710945977105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,65536,0.13656532764434814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,1024,0.006994666324721442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,16384,0.02387822171052297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,51200,0.058548443847232394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,51200,0.10689244667689006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,12288,0.01921066641807556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,65536,0.203313774532742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,512,0.006151111175616582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,16384,0.037422223223580256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,16384,0.021388444635603163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,10240,0.016520000166363187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,51200,0.15967378351423475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,12288,0.01699555582470364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,16384,0.05860888957977295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,8192,0.014135110709402295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,12288,0.029891557163662378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,10240,0.015338665909237333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,7168,0.0125591109196345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,10240,0.025647999511824712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,8192,0.012822222378518848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,12288,0.046068443192376025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,6144,0.011331555744012197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,10240,0.04035288757748074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,7168,0.011791110866599612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,7168,0.019330667124854196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,5120,0.009876444107956356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,8192,0.03432710965474447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,6144,0.0174026671383116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,7168,0.03126044405831231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,4096,0.00817599975400501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,5120,0.010311110979980892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,5120,0.01517599986659156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,4096,0.009377777576446533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,5120,0.024485333098305598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,4096,0.021376889612939622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,3584,0.012166221936543783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,3072,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,3584,0.009116444322797988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,3584,0.01963644391960568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,8192,0.021506667137145996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,2560,0.006096000058783426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,3072,0.011140444212489657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,3072,0.00869422240389718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,6144,0.01107377807299296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,2048,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,3072,0.01813066667980618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,6144,0.027220444546805486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,2560,0.016192888220151264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,1536,0.004592888885074192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,2048,0.009011555876996782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,3584,0.007548444800906711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,2048,0.015478221906556023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,1536,0.0074337778819931885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,1536,0.00795199970404307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,1536,0.013439999686347114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,1024,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,768,0.0031946665710873077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,1024,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,1024,0.011744889120260874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,2560,0.008344000412358178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,768,0.006911110960774952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,512,0.0028568889117903183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,768,0.011057777537239922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,512,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,256,0.00250133333934678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,2048,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,512,0.006337777607970768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,512,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,128,0.002467555511328909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,1024,0.003590222034189436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,256,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,256,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,64,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,128,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,256,0.010029333333174387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,32,0.002528888897763358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,128,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,65536,0.06491644514931573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,64,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,128,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,32,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,768,0.0063475556671619415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,65536,0.09640444649590386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,51200,0.04808177881770664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,65536,0.07134933604134454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,2560,0.01034311122364468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,16384,0.019156444403860304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,51200,0.056051552295684814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,51200,0.07617600096596612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,65536,0.2011786699295044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,4096,0.013019555144839816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,12288,0.016474665866957772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,16384,0.020623111062579684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,10240,0.01572088897228241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,51200,0.15971110926734075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,12288,0.016543999314308167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,12288,0.022343110707071092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,8192,0.013198222551080914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,10240,0.015169777803950839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,10240,0.01924088928434584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,12288,0.04558311237229241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,7168,0.011796444654464722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,8192,0.01276533305644989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,10240,0.0398613346947564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,7168,0.014947555131382413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,7168,0.011753777662913004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,8192,0.033017777734332614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,7168,0.03079644507831997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,6144,0.013771555489963956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,5120,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,6144,0.011029333704047732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,6144,0.02699911097685496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,4096,0.0075297777851422625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,16384,0.05784444676505195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,5120,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,4096,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,5120,0.023933332827356126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,4096,0.010676444404655032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,3584,0.0070186663005087115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,3584,0.009012444151772393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,8192,0.01632000009218852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,4096,0.020925333102544148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,3072,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,3584,0.010038221875826517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,2560,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,6144,0.010485333700974783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,3584,0.019640889432695176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,3072,0.009372444616423713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,2048,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,2560,0.008708444734414419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,3072,0.01756888959142897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,16384,0.02780710988574558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,1536,0.004197333421972063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,2560,0.008309333688682979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,2048,0.007777777810891469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,2048,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,2560,0.01644977761639489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,1024,0.0035128887328836652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,5120,0.012060444388124677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,1536,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,2048,0.014826666977670459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,768,0.003208000005947219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,1536,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,1024,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,1024,0.006315555423498154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,512,0.0028373334142896864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,768,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,1024,0.012044444680213928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,768,0.005966222120655908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,256,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,512,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,512,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,768,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,128,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,3072,0.008656889200210571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,256,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,512,0.011029333704047732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,64,0.0021484444538752237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,256,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,768,32,0.0021484444538752237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,128,0.006224888894293044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,65536,0.047741333643595375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,128,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,256,0.009480888644854227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,51200,0.03901333279079861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,64,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,128,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,65536,0.07112977902094524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,65536,0.09444089068306817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,16384,0.0177511109246148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,51200,0.055494222376081676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,51200,0.0747671127319336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,12288,0.013977777626779346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,65536,0.2004542218314277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,16384,0.019985778464211356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,16384,0.0270462218258116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,1536,0.013449778159459433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,51200,0.15927110777960882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,12288,0.016233777006467182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,12288,0.021433777279324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,16384,0.05749866697523329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,8192,0.012991111311647626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,10240,0.014875556031862894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,10240,0.019083556201722886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,12288,0.04510133465131124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,7168,0.011404444773991903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,8192,0.012646222280131446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,10240,0.038562665383021034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,7168,0.011708444191349877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,32,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,7168,0.015053333507643806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,8192,0.03273244367705451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,6144,0.009836444424258338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,5120,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,6144,0.01074222226937612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,7168,0.029482665989134047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,6144,0.013528889252079858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,4096,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,5120,0.011997333003415002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,5120,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,6144,0.026386666629049513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,10240,0.012730666332774691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,3584,0.006711111300521427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,4096,0.009327111144860586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,4096,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,5120,0.023731556203630235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,3072,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,3584,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,3584,0.00979644474056032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,2560,0.005535111245181825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,4096,0.020802666743596394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,3072,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,3072,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,8192,0.016228444046444364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,2048,0.004862222406599256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,3072,0.017557332913080852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,2560,0.008606221940782335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,1536,0.0041804446114434134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,2048,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,2560,0.016156444946924847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,2048,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,1024,0.003559999995761447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,2048,0.014763555592960782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,1536,0.006863111009200414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,1024,0.006303999986913469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,1024,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,768,0.0032266666077905228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,1536,0.013423110875818463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,1024,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,768,0.006647999915811751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,768,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,512,0.00630044440428416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,3584,0.019277332557572257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,256,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,512,0.0058417779703934984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,2560,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,128,0.0023013332651721104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,512,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,256,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,256,0.00999022192425198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,128,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,1536,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,64,0.0021439999755885866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,32,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,65536,0.03054844339688619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,128,0.007442666424645319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,64,0.005022222383154763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,32,0.005195555587609609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,51200,0.027738667196697656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,65536,0.09318488836288452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,512,512,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,65536,0.07058666812049018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,16384,0.012244444754388599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,768,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,51200,0.05464799867735969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,51200,0.07321155733532377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,65536,0.1995635562472873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,12288,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,16384,0.02643555568324195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,16384,0.019496889577971566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,256,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,10240,0.009418666362762451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,51200,0.15905600123935276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,12288,0.02088177700837453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,12288,0.016383111476898193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,8192,0.008244444098737504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,128,0.006098666538794835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,10240,0.014951111541854011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,10240,0.018271999226676095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,12288,0.04384266667895847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,7168,0.007718222008811102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,8192,0.0123848890264829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,8192,0.01573244399494595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,6144,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,10240,0.037754668129815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,7168,0.011733333269755045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,5120,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,7168,0.014478221535682678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,8192,0.03201155530081855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,6144,0.010743111371994019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,6144,0.013311111264758639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,7168,0.029153777493370905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,4096,0.006911110960774952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,5120,0.010000000397364298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,5120,0.011752888560295105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,3584,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,6144,0.02606933315594991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,4096,0.009150221943855286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,5120,0.023520888553725347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,4096,0.010314666562610203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,3072,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,3584,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,3584,0.009670221971140968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,4096,0.020374221934212577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,3072,0.008575110799736446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,3072,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,16384,0.05634755558437771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,2048,0.004599110947714912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,2560,0.008454221818182204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,2560,0.008151111503442129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,3072,0.017589333984586928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,1536,0.003871110992299186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,2048,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,2048,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,2560,0.015803555647532146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,1024,0.0032239999208185407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,1536,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,1536,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,2048,0.014594667487674289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,768,0.002862222285734283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,1024,0.006990222467316522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,1536,0.013064888616402945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,768,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,768,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,1024,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,512,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,2560,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,256,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,512,0.006303999986913469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,768,0.011061333119869232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,512,0.005531555662552516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,128,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,256,0.006199999815887875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,256,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,512,0.01072799993885888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,64,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,128,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,128,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,256,32,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,256,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,65536,0.02536711096763611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,64,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,128,0.007359111474619971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,51200,0.019687111179033916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,1024,0.006007110906971826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,65536,0.09295822514428033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,65536,0.06956533590952556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,16384,0.011368888947698804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,51200,0.05288889010747274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,51200,0.07317155599594116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,65536,0.20129777325524223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,12288,0.009396444592210982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,16384,0.026029333472251892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,16384,0.019481778144836426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,10240,0.008627555436558193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,51200,0.15808622042338052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,12288,0.016238222519556682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,16384,0.05572799841562907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,12288,0.020933333370420668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,8192,0.00776711106300354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,10240,0.014881778094503613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,10240,0.01822933389080895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,7168,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,8192,0.015503999259736804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,8192,0.012458666331238218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,10240,0.03769155674510532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,6144,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,8192,0.03224533465173509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,7168,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,32,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,7168,0.014411555396185981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,5120,0.0063662222690052455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,6144,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,6144,0.013029333617952136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,4096,0.006314666734801398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,7168,0.029057777590221826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,5120,0.011511999699804517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,5120,0.01001422190003925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,3584,0.006221333311663733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,6144,0.02607466611597273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,3584,0.019105777144432068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,4096,0.00924355536699295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,5120,0.02329866588115692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,4096,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,3072,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,3584,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,3584,0.009687999884287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,2560,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,4096,0.02051733268631829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,3072,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,12288,0.04376444551679823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,3584,0.018921777606010437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,2560,0.008060444560315874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,3072,0.01700177788734436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,2560,0.015842666228612263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,2048,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,1536,0.003823111040724648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,2048,0.014765333798196582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,1536,0.006765333314736684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,1024,0.0032115555885765287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,1536,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,1536,0.012774222426944308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,1024,0.006899555524190267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,1024,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,768,0.002862222285734283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,3072,0.00853066643079122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,1024,0.012044444680213928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,512,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,2048,0.004287999951177173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,768,0.005780444376998478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,2560,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,256,0.002159111201763153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,512,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,2048,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,512,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,128,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,256,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,512,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,128,0.005980444451173146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,256,0.009755555954244401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,64,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,128,32,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,128,0.00741155528359943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,65536,0.02402044501569536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,64,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,32,0.004971555537647671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,51200,0.018070222602950204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,16384,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,768,0.006647111227114995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,12288,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,10240,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,65536,0.09289511044820149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,51200,0.07278222507900663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,8192,0.006052444378534953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,16384,0.025560888979170058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,768,0.010784889260927836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,6144,0.005990222096443176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,12288,0.020635555187861126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,5120,0.006787555499209299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,10240,0.018031999468803406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,256,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,8192,0.01551822159025404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,4096,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,7168,0.014226666755146451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,3584,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,3072,0.005426666802830166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,2560,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,128,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,2048,0.004214222232500712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,5120,0.011534222298198275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,1536,0.0035546666218174826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,4096,0.010064889159467485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,3584,0.009408888717492422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,1024,0.0031884445084465873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,3072,0.008789333204428354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,768,0.0028204443968004654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,2560,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,2048,0.0075786668393347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,256,0.0022177776942650476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,1024,0.005987555616431766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,768,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,128,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,64,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,256,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,32,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,128,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,7168,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,65536,0.0184906671444575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,51200,0.03117777903874715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,32,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,16384,0.013067555924256643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,6144,0.013062222136391534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,12288,0.008612444831265343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,51200,0.07237866852018568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,10240,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,16384,0.025416000021828547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,64,512,0.0025048889219760895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,8192,0.006568000134494569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,7168,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,10240,0.018051554759343464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,1536,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,6144,0.006005333529578314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,5120,0.00565422202150027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,512,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,8192,0.01549600064754486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,4096,0.006171555568774541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,7168,0.01421511173248291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,3584,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,3072,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,6144,0.0129457778400845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,5120,0.011765333513418833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,2560,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,4096,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,3584,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,3072,0.008974221845467886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,1536,0.0038862224254343244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,2560,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,65536,0.09274666839175755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,1024,0.003216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,768,0.00282577777074443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,1536,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,512,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,1024,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,256,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,12288,0.02055644492308299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,128,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,512,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,64,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,256,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,32,0.0019279999865425958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,128,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,64,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,64,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,768,32,2048,0.004193777839342753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,2048,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,768,0.00565155554148886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,16384,0.7501137521531848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,16384,0.3125271002451579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,12288,0.524976889292399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,16384,0.7473031149970161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,32,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,12288,0.5739022360907661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,10240,0.4297395547231038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,12288,0.23659910096062553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,10240,0.2005502250459459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,10240,0.4536568853590224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,8192,0.4152924484676785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,51200,2.388743930392795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,51200,1.0054889255099826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,8192,0.16177599959903294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,7168,0.32511289914449054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,8192,0.3934577835930719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,6144,0.27435289488898384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,5120,0.25662488407558864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,7168,0.1460168891482883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,7168,0.31742223103841144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,6144,0.12530755996704102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,51200,2.562699635823568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,4096,0.20751467016008165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,3584,0.19394844108157686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,6144,0.39439998732672793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,4096,0.08773777882258098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,5120,0.10488977697160508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,4096,0.19815644952985975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,3584,0.16336622503068712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,5120,0.23546489079793295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,3584,0.07610044214460585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,2560,0.12812977366977268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,2048,0.10130133231480916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,3072,0.0683715542157491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,3072,0.16394133037990993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,1536,0.08438399765226577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,2560,0.05903822183609009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,1024,0.05645066499710083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,2048,0.052411556243896484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,1536,0.04408177733421326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,768,0.04273688793182373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,2048,0.10391910870869954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,1536,0.07961777846018474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,1024,0.03270577722125583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,512,0.029956443442238703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,256,0.02259644369284312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,768,0.028856890069113836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,512,0.02688088847531213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,768,0.04885955651601156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,1024,0.060602664947509766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,128,0.01902933253182305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,256,0.023966223001480103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,64,0.01645511057641771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,256,0.03739733497301737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,32,0.016919111212094624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,128,0.023418666587935552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,128,0.03495555453830295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,3072,0.1526177724202474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,64,0.03545777665244208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,32,0.03460266523891025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,2560,0.12242044342888726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,512,0.04234400060441759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,16384,0.5772550900777181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,16384,0.2444302241007487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,16384,0.5798817740546333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,51200,1.913900375366211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,65536,1.0329368379380968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,65536,2.4522266387939453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,51200,1.9000657399495442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,12288,0.47681511773003477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,65536,2.501720004611545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,12288,0.4275982115003798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,8192,0.32460443178812665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,10240,0.44863467746310765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,10240,0.1608417828877767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,12288,0.1879875527487861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,8192,0.1284995608859592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,7168,0.2698977788289388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,10240,0.37048178248935276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,7168,0.1147528886795044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,51200,0.7821057637532552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,7168,0.25054489241706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,6144,0.20807288752661812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,5120,0.2053582270940145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,6144,0.09823377927144368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,5120,0.08433333370420669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,6144,0.227729770872328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,5120,0.18205066521962485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,3584,0.14964266618092856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,4096,0.0690808892250061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,4096,0.14883111582862005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,3072,0.10838844378789265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,3584,0.06126755475997925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,2560,0.09806311130523682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,3584,0.12962133354610866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,3072,0.11220888296763103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,2048,0.0981031126446194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,2560,0.0477128889825609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,8192,0.2862897713979085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,2560,0.09610666831334432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,2048,0.04074133435885111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,1024,0.051515556044048734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,1536,0.03245066603024801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,1536,0.06365955538219877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,1024,0.02549422283967336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,1024,0.04881422387229072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,768,0.02306844459639655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,512,0.027088888817363318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,768,0.03933777742915683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,3072,0.05490044421619839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,512,0.021115556359291077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,256,0.01920622256067064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,512,0.033097777101728655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,128,0.014766222900814481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,256,0.01939644416173299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,1536,0.07364888985951741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,256,0.029120888974931505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,128,0.01925244430700938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,32,0.01332266628742218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,2048,0.08035377661387126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,128,0.028104000621371802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,4096,0.14575821823543972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,64,0.02812444501452976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,768,0.03447733322779337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,65536,0.7272266811794705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,51200,0.5521368980407715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,65536,0.3404257827334934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,51200,0.26827822791205513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,16384,0.2096560001373291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,65536,0.7927298016018338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,51200,0.6232284439934624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,12288,0.1673964394463433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,16384,0.0996204415957133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,10240,0.11802666717105442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,16384,0.200546662012736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,8192,0.0970497793621487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,12288,0.1530417733722263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,10240,0.05896266963746813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,32,0.027689778142505225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,10240,0.12645955880482992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,7168,0.09283377726872762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,8192,0.09887199931674534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,8192,0.046431110964881055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,7168,0.04150933358404372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,5120,0.06318933433956571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,7168,0.08824621968799168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,6144,0.03784977727466159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,4096,0.05223911007245382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,6144,0.07767289214664035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,5120,0.03138577938079834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,3584,0.04786044359207153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,5120,0.06418133444256253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,3072,0.04216177927123176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,4096,0.05345333284801907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,3584,0.02417244513829549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,2560,0.03568533394071791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,64,0.013067555924256643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,3584,0.04685689012209574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,2048,0.028232001596026953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,3072,0.04122044311629401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,2560,0.019541333119074505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,3072,0.021912000245518152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,1536,0.02308799988693661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,2560,0.035812444157070585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,2048,0.017493334081437852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,1024,0.015751111838552687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,2048,0.029750221305423315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,1536,0.015476443701320224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,768,0.012472888661755456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,6144,0.07778577672110663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,1024,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,512,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,1024,0.019499555230140686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,768,0.011618666350841522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,256,0.007376000285148621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,768,0.016547555724779766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,512,0.010760000182522668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,128,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,512,0.013796444568369122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,256,0.010004444254769219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,256,0.012712000144852532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,64,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,4096,0.027504000398847792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,128,0.010075555907355415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,32,0.006081777728266186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,64,0.012094222009181976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,128,0.012238221863905588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,32,0.012103111379676394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,12288,0.06673333379957411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,65536,0.47617509629991317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,51200,0.35390400886535645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,16384,0.1352355612648858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,1536,0.024678222007221643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,65536,0.24905244509379068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,65536,0.5966053538852268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,12288,0.08851199679904515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,51200,0.47974220911661786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,51200,0.19360266791449654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,10240,0.0826524429851108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,12288,0.05018933282958137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,16384,0.07062221897972955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,8192,0.07427289088567098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,16384,0.1515173382229275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,12288,0.1174737744861179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,7168,0.05888622336917453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,10240,0.04277955492337545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,8192,0.03660533494419522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,6144,0.050719112157821655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,7168,0.0314408904976315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,10240,0.0931733316845364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,8192,0.07558311356438531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,5120,0.046739555067486234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,7168,0.06919555531607734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,4096,0.042668445242775806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,6144,0.029273778200149536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,3584,0.03211999932924906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,5120,0.024336000283559162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,3072,0.031537777847713895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,4096,0.02347822156217363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,5120,0.05145155390103658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,3584,0.019095111224386428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,2560,0.02676533328162299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,3584,0.03715377714898851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,3072,0.018046221799320646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,2048,0.02167111138502757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,2560,0.01633777717749278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,3072,0.032871110571755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,1536,0.016232000456915963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,1024,0.011959111524952782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,2560,0.028604444530275133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,2048,0.014399111270904541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,1536,0.012778666284349231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,768,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,2048,0.024316444993019104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,1536,0.020071110791630216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,1024,0.010696888797812991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,768,0.009861333502663506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,256,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,512,0.009331555830107795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,768,0.013762666947311826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,128,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,512,0.011360888679822287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,256,0.008647110727098253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,64,0.0042408886882993905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,256,0.01074399964676963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,128,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,6144,0.05970399909549289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,128,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,4096,0.04180888997183906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,64,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,32,0.010336000058386061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,65536,0.5173306465148926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,512,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,51200,0.4138302273220486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,65536,0.5124239921569824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,1024,0.016199999385409884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,51200,0.4073048962487115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,16384,0.13205156061384413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,51200,0.1781768931282891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,16384,0.06190133094787598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,10240,0.09557511409123738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,16384,0.12361955642700195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,32,0.004570666700601578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,12288,0.09398755762312148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,8192,0.07447199689017402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,7168,0.07067022058698866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,10240,0.0411155554983351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,8192,0.03593688872125413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,10240,0.07955200142330594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,6144,0.06042844719356961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,7168,0.03141511148876614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,7168,0.05820000171661377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,5120,0.049867557154761426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,6144,0.026379555463790894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,6144,0.051023112403021924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,4096,0.037853333685133196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,5120,0.023278221487998962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,65536,0.22867290178934732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,5120,0.04333688815434774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,3584,0.0347688893477122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,4096,0.020289777053727042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,3072,0.030638221237394545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,12288,0.10112622049119736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,3584,0.018588445252842374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,2560,0.025224000215530396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,3584,0.03200444579124451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,12288,0.04661688870853848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,3072,0.01704266667366028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,2048,0.020787555310461257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,3072,0.028749333487616643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,2560,0.01534577707449595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,2048,0.01401600076092614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,2560,0.02494488822089301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,1024,0.012359110845459832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,2048,0.02141777839925554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,8192,0.06487022505866157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,1536,0.011801777614487542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,768,0.010249777800507015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,1536,0.01791288952032725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,512,0.007836444510353936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,1024,0.010280888941552904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,1024,0.014360000689824423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,768,0.009663999908500248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,256,0.006344888773229387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,512,0.009009777671760982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,768,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,512,0.01036088913679123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,256,0.008626666333940294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,64,0.004224888980388641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,256,0.009959999885823991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,32,0.004552000098758274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,4096,0.03572533196873135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,128,0.00943199959066179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,64,0.009345778160625035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,32,0.009572444690598382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,1536,0.016362667083740234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,65536,0.3364124563005235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,51200,0.28953689999050564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,65536,0.41727733612060547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,65536,0.20220800240834555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,16384,0.09506755405002171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,51200,0.14179200596279568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,12288,0.06674489047792223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,51200,0.3285040060679118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,16384,0.051206221183141075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,10240,0.06183822287453545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,16384,0.1023484468460083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,8192,0.05025066600905525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,12288,0.039220445685916476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,10240,0.03407288922203912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,12288,0.07803733481301202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,128,0.008635555704434713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,7168,0.04749777913093567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,10240,0.06592888964547051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,6144,0.04234133495224846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,8192,0.027002667387326557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,5120,0.034859554635153875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,7168,0.024626667300860088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,8192,0.05442488855785794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,4096,0.02790400054719713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,6144,0.022035555707083807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,7168,0.04855022165510389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,3584,0.024668445189793903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,6144,0.0432933337158627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,3072,0.021738666627142165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,4096,0.017515555024147034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,5120,0.036624001132117376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,3584,0.015863110621770222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,2560,0.018213333355055917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,3584,0.027394667267799377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,2048,0.01519733336236742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,3072,0.024340444140964087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,2560,0.013442666994200813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,1536,0.012563555604881711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,2560,0.021418665846188862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,2048,0.012063999970753988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,128,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,2048,0.018339556124475267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,1536,0.010646222366227044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,768,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,1536,0.01552888916598426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,1024,0.009178666604889764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,512,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,768,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,1024,0.012077333198653327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,256,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,768,0.010685332947307162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,512,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,128,0.004210666649871402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,512,0.009338666995366415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,256,0.007821333077218797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,5120,0.019575110740131803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,64,0.0038453332251972626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,256,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,32,0.0038986665507157645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,128,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,4096,0.030048890246285334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,128,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,3072,0.014816888504558139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,64,0.008636444807052612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,32,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,65536,0.307948456870185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,51200,0.24134310086568198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,1024,0.009782222410043081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,16384,0.08282666736178927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,65536,0.1700648864110311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,12288,0.06682844294442071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,65536,0.36901066038343644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,51200,0.2843315601348877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,16384,0.04659822252061632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,10240,0.051048888100518115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,16384,0.09120444456736247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,12288,0.037831111086739436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,8192,0.0473599996831682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,7168,0.0380311111609141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,10240,0.05924266576766968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,8192,0.026802667313151892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,6144,0.037272890408833824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,8192,0.0483822226524353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,7168,0.023762666516833838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,7168,0.043215112553702466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,5120,0.02903911140229967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,6144,0.02200355629126231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,6144,0.038423998488320246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,4096,0.024677332904603746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,5120,0.019261333677503798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,3584,0.021382222572962444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,5120,0.03262133399645487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,4096,0.017071111334694754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,3072,0.01848444508181678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,4096,0.02755111124780443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,3584,0.015650666422314115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,2560,0.01590488851070404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,51200,0.13531733883751765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,3072,0.014336888988812765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,3584,0.024851555625597637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,2048,0.01423377792040507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,3072,0.022459555003378127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,2560,0.01310222182008955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,2048,0.011714666253990598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,2560,0.019533332851197984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,12288,0.06997422377268474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,1024,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,2048,0.016897777716318767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,1536,0.009927999642160203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,10240,0.033040000332726374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,768,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,512,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,1536,0.01419911119672987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,1024,0.009012444151772393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,256,0.004557333472702238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,768,0.008381333616044786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,1024,0.011163555913501315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,128,0.003887999802827835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,512,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,768,0.010044444766309526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,64,0.003547555456558863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,256,0.00778399987353219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,32,0.003536888708670934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,128,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,256,0.008395555946562026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,128,0.00815288888083564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,64,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,32,0.008151111503442129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,51200,0.18528000513712564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,65536,0.13272799385918513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,16384,0.06033866935306125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,65536,0.32322488890753853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,51200,0.1049235529369778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,1536,0.011612444288200803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,12288,0.04773866799142626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,51200,0.254896879196167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,10240,0.038478222158220075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,16384,0.08054488897323608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,12288,0.028970665401882593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,8192,0.03752266698413425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,12288,0.06151555644141304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,7168,0.029324445459577773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,8192,0.02122488949033949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,10240,0.052410665485594005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,6144,0.024519110719362896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,8192,0.04297955499755012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,7168,0.02295733326011234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,5120,0.024130667249361675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,6144,0.017874666386180453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,7168,0.03809066613515218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,4096,0.020004444652133517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,6144,0.03382755650414361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,65536,0.2587920029958089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,5120,0.016121778223249648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,3584,0.017303110824690927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,5120,0.029550221231248643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,3072,0.015088889333936902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,4096,0.024907555845048692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,3584,0.01310488912794325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,3584,0.022456000248591106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,2560,0.013188444077968597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,3072,0.012019555601808759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,3072,0.020277332928445604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,2048,0.011038222246699862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,2560,0.010703110860453712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,2560,0.01778133379088508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,1536,0.00924622184700436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,10240,0.024999999337726172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,2048,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,2048,0.015224888920783997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,1024,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,1536,0.00963644435008367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,768,0.006221333311663733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,1536,0.01255022237698237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,1024,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,512,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,1024,0.010399999717871347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,768,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,256,0.0038373333712418876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,768,0.009331555830107795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,512,0.0075937774446275495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,4096,0.015956444872750174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,128,0.0032008888406885993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,512,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,64,0.00315022220214208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,256,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,32,0.0031884445084465873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,256,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,128,0.00703288863102595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,128,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,16384,0.038098666403028704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,64,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,32,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,65536,0.2939751148223877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,51200,0.20292354954613578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,16384,0.07526310947206286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,65536,0.12342845069037543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,65536,0.27448799875047475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,512,0.008715555899673039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,12288,0.055027554432551064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,51200,0.0982053346104092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,16384,0.039022223816977605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,10240,0.046726223495271474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,8192,0.036905778778923884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,12288,0.02734666731622484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,16384,0.06872000296910603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,12288,0.05336088935534159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,10240,0.025937777426507738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,7168,0.03243466549449497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,6144,0.02796622117360433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,8192,0.02029244436158074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,7168,0.022216889593336318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,8192,0.037456889947255455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,5120,0.02424000038041009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,6144,0.017295110556814406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,4096,0.019811555743217468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,5120,0.015850666496488783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,4096,0.014723555909262763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,3584,0.01759111053413815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,5120,0.02584533393383026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,3072,0.015378667248619927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,3584,0.012729778057999082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,3584,0.020021332634819877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,2560,0.01350844403107961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,3072,0.011757333245542316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,3072,0.018053332964579265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,2048,0.011271110839313932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,51200,0.21316799852583143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,2560,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,1536,0.009551111194822524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,2560,0.01585244470172458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,2048,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,2048,0.013204444613721637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,1024,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,10240,0.04555111130078634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,1536,0.009014222357008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,768,0.005987555616431766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,1536,0.011361777782440186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,1024,0.008230222596062554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,512,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,7168,0.03333244389957852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,768,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,1024,0.009485333330101436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,768,0.008504889077610439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,6144,0.029612445169025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,128,0.003555555724435382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,256,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,512,0.007678222325113084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,64,0.0032008888406885993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,4096,0.02201066745652093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,32,0.0033075554917256036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,128,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,64,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,65536,0.21036177211337617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,32,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,51200,0.16484888394673666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,65536,0.11672800117068821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,65536,0.2206266721089681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,16384,0.0599493318133884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,51200,0.17628355820973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,51200,0.09346666600969102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,12288,0.04558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,16384,0.03351466523276435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,256,0.004228444563017951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,16384,0.058192001448737256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,512,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,10240,0.03938222262594435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,12288,0.02676533328162299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,8192,0.03177422285079956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,12288,0.04475466741455925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,10240,0.02388444377316369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,128,0.006991110742092133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,7168,0.02644799980852339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,8192,0.020113777783181932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,6144,0.02252888845072852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,10240,0.03805511196454366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,5120,0.019270223047998216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,7168,0.01917866700225406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,8192,0.03174044357405768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,4096,0.016075554821226332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,7168,0.02858577834235297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,5120,0.015058666467666626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,6144,0.02568622264597151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,3584,0.014800000521871777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,5120,0.02221333318286472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,4096,0.013064888616402945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,3072,0.012682666381200155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,4096,0.019165333774354722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,2560,0.011247999966144562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,3584,0.011932444241311816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,3584,0.01756088932355245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,3072,0.011196444431940714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,2048,0.009826666778988307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,3072,0.016137777103318107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,2560,0.010060444474220276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,1536,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,2560,0.013665777941544851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,1024,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,2048,0.011724444727102915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,1536,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,768,0.005434666656785541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,1536,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,512,0.004517333375083076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,256,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,1024,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,768,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,256,0.00351555562681622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,768,0.007781333393520779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,512,0.0071928890215026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,128,0.003504000190231535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,256,0.006913777854707506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,64,0.0028311111446883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,256,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,128,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,32,0.0029928889125585556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,128,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,64,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,65536,0.17932000425126818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,32,0.006595555692911148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,51200,0.14034577210744223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,2048,0.00923822240697013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,65536,0.21405243873596191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,16384,0.05268088976542155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,51200,0.09615377585093181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,51200,0.16593866878085667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,12288,0.040004445446862116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,16384,0.033193777004877724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,16384,0.05650399790869819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,10240,0.03408977720472548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,12288,0.0274008893304401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,512,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,8192,0.02737688852681054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,10240,0.023444443941116333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,7168,0.023784889115227595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,10240,0.037223110596338906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,8192,0.02033866610791948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,6144,0.02035199933581882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,8192,0.031141334109836157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,7168,0.01792355544037289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,5120,0.01666844387849172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,7168,0.028199109766218398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,6144,0.016484444340070088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,4096,0.0144177774588267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,5120,0.014479110638300577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,6144,0.025174222058720056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,3584,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,5120,0.021687999367713928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,4096,0.012778666284349231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,3072,0.011679111255539788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,4096,0.018813333577579923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,3584,0.011747555600272285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,3584,0.017085333665211994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,2560,0.00997688869635264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,3072,0.010743111371994019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,2048,0.008621333373917473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,3072,0.0151555554734336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,2560,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,1536,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,6144,0.016799110505315993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,2560,0.013053333593739403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,2048,0.009034666750166152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,1024,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,2048,0.011423110961914062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,1536,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,768,0.004861333303981357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,1536,0.010041777458455827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,1024,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,512,0.0038515557017591265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,1024,0.008637333081828224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,768,0.007330666813585494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,256,0.003261333331465721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,512,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,512,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,128,0.0028311111446883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,256,0.006695111178689533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,256,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,64,0.002569777684079276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,1024,0.007697777615653143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,128,0.00665155549844106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,32,0.0028488888508743713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,128,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,65536,0.18727644284566244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,12288,0.04310755597220527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,32,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,51200,0.12769867314232722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,65536,0.1801839934455024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,65536,0.0740248892042372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,16384,0.049419555399152965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,51200,0.14101155598958334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,51200,0.0616497794787089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,16384,0.022671111755900916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,12288,0.03641600079006619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,768,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,16384,0.04667199982537163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,65536,0.11545688576168484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,8192,0.024668445189793903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,10240,0.017601778109868366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,12288,0.0362426671716902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,10240,0.03142755561404758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,7168,0.020627554919984605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,8192,0.014152889450391134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,6144,0.018220444520314533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,8192,0.02622844444380866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,7168,0.023611555496851604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,64,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,5120,0.015491555134455362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,6144,0.012113778127564324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,4096,0.013380444712109037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,6144,0.021407999926143225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,3584,0.011988444460762871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,5120,0.011381333072980246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,5120,0.018862222631772358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,4096,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,3072,0.010799999866220685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,4096,0.01588000026014116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,3584,0.010669333239396414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,2560,0.009342222577995723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,3584,0.014467555615637036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,3072,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,2048,0.008062221937709385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,10240,0.030608889129426744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,3072,0.01314399970902337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,2560,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,1536,0.006996444529957241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,12288,0.018910222583346896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,2048,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,1536,0.007612444460391998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,2048,0.010048000348938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,1024,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,1024,0.007676444119877285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,512,0.0038444445365005066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,768,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,768,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,256,0.0032151111712058387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,512,0.006320888797442119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,512,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,128,0.0028115556471877625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,256,0.00630666646692488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,256,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,64,0.0025315555847353404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,128,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,128,0.006117333140638139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,32,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,64,0.006009777800904379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,1024,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,2560,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,32,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,768,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,1536,0.008991111483838823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,65536,0.07159644365310669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,16384,0.045328888628217906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,65536,0.16839910878075492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,51200,0.058485335773891874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,12288,0.035199112362331815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,16384,0.021312889125612047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,10240,0.02887111239963108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,16384,0.04529866576194763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,12288,0.018230222993426852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,12288,0.035338666703965925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,10240,0.016508445143699646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,8192,0.02346311178472307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,10240,0.030525333351559106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,7168,0.020845333735148113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,8192,0.013835555149449242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,8192,0.025811556312772963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,6144,0.018550222118695576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,65536,0.17482844988505045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,7168,0.023240889112154644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,5120,0.015778667396969266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,6144,0.01166311071978675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,51200,0.13023555278778076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,6144,0.020977778567208186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,4096,0.013259555730554791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,5120,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,7168,0.013760888742076026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,3584,0.012092444631788464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,51200,0.133607996834649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,4096,0.009323555562231276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,3072,0.010968888799349466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,4096,0.015327110886573792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,3584,0.00906044410334693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,2560,0.009672889278994666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,3072,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,2048,0.008273777862389883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,2560,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,2560,0.011156444748242697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,2048,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,2048,0.009806222385830348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,7168,0.01275466630856196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,1024,0.00518133325709237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,1536,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,1536,0.008749333520730337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,768,0.004601777841647466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,1024,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,1024,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,5120,0.018211555149820115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,768,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,512,0.003526222374704149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,768,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,256,0.00282222218811512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,512,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,256,0.006100444330109491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,128,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,512,0.006540444576077991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,256,0.006082666830884085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,64,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,3072,0.01273333364062839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,32,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,128,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,128,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,1536,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,65536,0.1303280062145657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,64,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,32,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,65536,0.0679360032081604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,16384,0.037632889217800565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,12288,0.03084533413251241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,51200,0.054670221275753446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,16384,0.020959999826219346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,51200,0.10233510865105523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,10240,0.027706666125191584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,12288,0.017424000634087455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,8192,0.021930666433440313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,12288,0.028170665105183918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,10240,0.015821332732836407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,7168,0.019903999235894945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,8192,0.013110222087966071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,8192,0.020810667011472914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,6144,0.017824000782436795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,7168,0.012053333222866058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,3584,0.014129777749379476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,7168,0.018960888187090557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,5120,0.014485332700941296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,6144,0.011076444553004371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,4096,0.01088177743885252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,6144,0.017222222354676988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,5120,0.015096889601813423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,3584,0.009958222508430481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,4096,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,4096,0.0127368892232577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,65536,0.13085156016879612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,3072,0.008969777988062965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,3584,0.008987555901209513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,2560,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,3584,0.011764444410800934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,3072,0.00850844466023975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,2048,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,3072,0.010771555205186209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,2560,0.008081778056091731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,1536,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,10240,0.024432889289326135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,2560,0.009580444130632613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,2048,0.007833778030342525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,2048,0.008622222476535374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,1536,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,768,0.0038737778862317405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,1536,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,1024,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,768,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,5120,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,768,0.0063528890411059065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,512,0.0034471110751231513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,256,0.002838222309947014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,51200,0.09382577737172444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,512,0.006311999840868845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,128,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,512,0.005970666805903117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,256,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,256,0.00563733321097162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,64,0.0023004443695147834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,32,0.002475555572244856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,16384,0.03529422150717841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,128,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,128,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,65536,0.09674311346477932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,64,0.005630222045713001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,32,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,51200,0.0772835546069675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,65536,0.06873955329259236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,16384,0.030210667186313208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,65536,0.1295457813474867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,51200,0.05379199981689453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,1024,0.006681777950790193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,51200,0.10087200005849202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,16384,0.03461688756942749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,10240,0.021592888567182753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,12288,0.01716800034046173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,8192,0.017669333351982962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,12288,0.027600889404614765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,10240,0.015171556009186639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,7168,0.016062221593326993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,10240,0.024093333217832778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,8192,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,6144,0.014123555686738757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,8192,0.020297777321603563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,7168,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,5120,0.012229333321253458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,7168,0.018352889352374606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,6144,0.011017777853541903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,4096,0.009639110830095079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,6144,0.0162453336848153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,5120,0.010044444766309526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,3584,0.009469333622190688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,1024,0.004286222159862518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,5120,0.014529777897728814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,4096,0.009126221968068017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,3072,0.007682667010360294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,4096,0.012440889245933957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,2560,0.007032000356250339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,3584,0.008656889200210571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,12288,0.02496799992190467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,3072,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,2048,0.0058151111006736755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,16384,0.02053155501683553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,3584,0.011575111084514193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,1536,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,3072,0.010601777997281818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,1024,0.003951111187537511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,2048,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,2560,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,768,0.0035013332962989807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,1536,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,2048,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,512,0.003022222262289789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,1024,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,1024,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,768,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,256,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,768,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,512,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,128,0.002481777841846148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,512,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,256,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,32,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,256,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,128,0.005333333379692501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,65536,0.057769775390625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,64,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,32,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,51200,0.052460445298088916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,65536,0.0654968884256151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,65536,0.08232000139024523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,16384,0.01977599991692437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,51200,0.05337511168585884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,2560,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,12288,0.016629333297411602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,51200,0.06525422467125787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,10240,0.015784000356992085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,16384,0.01994755533006456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,16384,0.024325332707828943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,1536,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,8192,0.013598221871587964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,10240,0.014480888843536377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,12288,0.019606222709019978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,7168,0.011985777980751462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,10240,0.01719555589887831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,8192,0.012443555725945367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,6144,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,64,0.002415111081467734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,7168,0.013409777647919126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,128,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,6144,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,6144,0.012448000411192576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,4096,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,5120,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,5120,0.011056889262464313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,3584,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,4096,0.009029332962301042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,4096,0.010016000105275048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,3072,0.006360000206364526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,3584,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,3584,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,2560,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,3072,0.008296888735559251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,3072,0.008758222063382467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,12288,0.01608000033431583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,2048,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,2560,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,1536,0.00425866660144594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,2560,0.007914666500356462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,8192,0.014518222875065274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,1024,0.003526222374704149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,1536,0.007147555549939473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,5120,0.009304888546466827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,768,0.0031617778456873367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,1536,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,1024,0.006656888872385025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,512,0.002871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,1024,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,768,0.006679111056857639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,256,0.002461333241727617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,768,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,128,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,512,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,256,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,512,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,64,0.0021235555824306277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,32,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,256,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,65536,0.05317155520121256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,64,0.005105777747101254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,32,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,51200,0.04237066706021627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,2048,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,65536,0.06455910868114896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,7168,0.01163200040658315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,16384,0.016404444972674053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,51200,0.05096711052788628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,16384,0.0196435550848643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,51200,0.0628604425324334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,16384,0.023623999622133043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,10240,0.01276622215906779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,12288,0.015827554795477126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,12288,0.018768888380792405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,10240,0.01442488862408532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,8192,0.013744889034165276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,10240,0.016560888952679105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,128,0.005992888990375731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,7168,0.012721777790122561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,8192,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,128,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,8192,0.014114666316244336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,7168,0.011459555890825061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,5120,0.009633777870072259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,7168,0.013007111019558378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,6144,0.010686222049925061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,4096,0.00742222203148736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,6144,0.01205244412024816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,65536,0.07966844240824382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,2048,0.007232888705200619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,5120,0.010861333045694562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,4096,0.008993777963850234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,3072,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,4096,0.009639999932712978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,3584,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,2560,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,3584,0.009249777429633671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,3072,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,2048,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,3072,0.008464888566070134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,2560,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,1536,0.004218666503826777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,2560,0.007711999946170383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,2048,0.00739644467830658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,1024,0.0035111109415690103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,2048,0.0070568886068132185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,1536,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,768,0.003157333367400699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,6144,0.010436444646782346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,1536,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,512,0.0028693332440323303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,1024,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,1024,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,768,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,256,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,512,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,768,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,128,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,256,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,5120,0.010031999813185798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,64,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,256,0.0053102220926019884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,32,0.0022124445272816554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,128,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,128,0.004993777722120285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,65536,0.03874489002757602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,64,0.005041777673694822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,32,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,65536,0.06368888749016656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,65536,0.07829955551359388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,16384,0.01311288856797748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,51200,0.049944000111685864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,51200,0.06103999747170342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,12288,0.013103110922707452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,16384,0.018648000227080453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,16384,0.02296977738539378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,10240,0.011794666449228922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,12288,0.01568177839120229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,12288,0.018294221825069852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,512,0.005449777675999536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,8192,0.010728889041476779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,10240,0.01443199978934394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,10240,0.016361777981122334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,7168,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,8192,0.012120889292822944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,6144,0.009577777650621202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,7168,0.011363555159833698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,8192,0.014121777481502958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,5120,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,7168,0.012759110993809171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,4096,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,51200,0.031077331966824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,6144,0.010701333483060202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,3584,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,5120,0.009731555150614845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,6144,0.011720889144473605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,3072,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,4096,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,5120,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,2560,0.005199111170238919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,3584,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,4096,0.00959022260374493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,2048,0.004504000147183736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,3584,0.008984000318580205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,3072,0.008105778031879002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,1536,0.003948444293604957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,3072,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,2048,0.00757244477669398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,2560,0.00758755538198683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,1024,0.003160888950030009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,1536,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,2048,0.0069208890199661255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,768,0.0028764444092909494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,1536,0.006595555692911148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,3584,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,512,0.002516444358560774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,1024,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,256,0.002479111154874166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,512,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,768,0.00565244464410676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,128,0.0021191111041439907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,512,0.0053688887920644544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,256,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,64,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,256,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,128,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,512,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,128,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,65536,0.02436711059676276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,64,0.004980444494220945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,32,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,51200,0.01963555481698778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,65536,0.07691466808319092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,2560,0.007752888732486301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,16384,0.010646222366227044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,51200,0.04837333493762546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,12288,0.010052444206343757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,51200,0.05934933159086439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,16384,0.018624888526068795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,10240,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,1024,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,16384,0.022145777940750122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,8192,0.00941422250535753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,768,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,10240,0.01425866617096795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,12288,0.017999110950364005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,7168,0.008646222452322641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,10240,0.01607733302646213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,6144,0.008112000094519721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,8192,0.013812444276279874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,7168,0.011367999845080905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,5120,0.007430222299363878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,7168,0.01276088919904497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,6144,0.010432889064153036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,4096,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,6144,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,5120,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,3584,0.0064204442832205035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,65536,0.06063555346594917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,5120,0.010447111394670276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,4096,0.008991111483838823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,3072,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,4096,0.009331555830107795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,3584,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,2560,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,3584,0.008966222405433655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,3072,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,2048,0.004446222136418025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,2560,0.007872888611422645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,12288,0.0155102229780621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,2560,0.0075528886583116316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,768,12288,0.013656888571050433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,2048,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,1024,0.003148444410827425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,8192,0.012101333174440594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,1536,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,1536,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,768,0.0028213332924577924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,1024,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,512,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,1024,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,768,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,256,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,768,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,512,0.006047111004590988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,128,0.0018195555441909367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,512,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,256,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,64,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,128,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,32,0.0018684443914228016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,128,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,65536,0.017299556069903903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,3072,0.008330666356616551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,64,0.005207111024194294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,51200,0.014263111684057446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,256,1536,0.003843555433882607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,65536,0.06006933583153618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,2048,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,65536,0.07610044214460585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,51200,0.0465280016263326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,12288,0.009492444495360056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,51200,0.05879288911819458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,16384,0.018560000591807894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,10240,0.008572444319725037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,16384,0.02203111184967889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,12288,0.01789422167672051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,8192,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,10240,0.014180445008807711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,10240,0.015841777125994366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,8192,0.012063110868136087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,256,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,8192,0.013639999760521783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,7168,0.011342222491900126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,7168,0.012443555725945367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,5120,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,6144,0.010398222340477837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,6144,0.011404444773991903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,4096,0.006911110960774952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,5120,0.009712888962692684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,16384,0.011120888921949597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,5120,0.010445333189434476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,3584,0.006322666588756774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,4096,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,4096,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,3072,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,3584,0.008584889272848764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,2560,0.0051893335249688895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,12288,0.01548266741964552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,3584,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,3072,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,2048,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,2560,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,3072,0.008353778057628209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,1536,0.0038533334930737815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,7168,0.008602666358153025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,2560,0.007544889218277401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,1024,0.003192000091075897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,2048,0.006859555426571105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,1536,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,6144,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,768,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,512,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,1536,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,1024,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,768,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,256,0.0021928888228204516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,512,0.005984888722499211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,768,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,128,0.0021342221233579847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,32,0.004954666727119022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,512,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,64,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,128,32,0.0022026666750510535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,128,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,256,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,65536,0.015160888433456421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,51200,0.013002666334311167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,128,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,64,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,16384,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,65536,0.07447733481725057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,12288,0.011361777782440186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,51200,0.05790044201744927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,10240,0.010070222119490305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,16384,0.021948445174429152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,2048,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,8192,0.008956444760163626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,7168,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,12288,0.017792888813548617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,6144,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,1024,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,10240,0.015796444482273526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,5120,0.006999111009968652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,8192,0.013598221871587964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,4096,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,7168,0.012460444536474017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,6144,0.011394666300879585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,5120,0.01035288886891471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,3072,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,2560,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,3584,0.00869066682126787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,2048,0.004230222354332606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,3072,0.008250666989220513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,32,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,1536,0.003567999849716822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,1024,0.0031795555518733132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,2048,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,768,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,1536,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,512,0.002510222295920054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,768,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,256,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,512,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,128,0.0019048889064126545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,3584,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,256,0.0058746664888328975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,64,32,0.001806222316291597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,4096,0.009268444445398118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,65536,0.035311109489864774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,64,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,51200,0.027775110469924078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,32,0.004860444615284602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,16384,0.014774221513006421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,2560,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,12288,0.020000888241661918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,10240,0.017312000195185345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,65536,0.07520711421966553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,8192,0.014901333385043673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,51200,0.0574586656358507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,16384,0.02180800007449256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,7168,0.013440888788965015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,12288,0.017735110388861764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,6144,0.012084444363911947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,5120,0.010427555276287926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,10240,0.015750222735934787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,4096,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,8192,0.013451555536852943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,7168,0.01237777786122428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,3584,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,6144,0.01143200033240848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,3072,0.006521777974234686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,2560,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,5120,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,2048,0.004238222208287981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,4096,0.009107555780145857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,1536,0.003565333369705412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,256,0.004956444518433677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,3584,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,1024,0.003100444459252887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,128,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,3072,0.008023999631404877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,512,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,2048,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,256,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,1536,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,128,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,1024,0.005886222339338726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,64,0.00179822225537565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,512,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,32,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,256,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,128,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,1024,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,64,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,32,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,513,32,768,0.0028479999552170434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,2560,0.007466666400432587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,16384,0.6049751175774468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,768,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,12288,0.47013335757785374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,16384,0.23275200525919595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,16384,0.6702853308783637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,51200,0.7423813078138563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,51200,2.164422141181098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,12288,0.1783991124894884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,12288,0.5081173578898112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,51200,2.250170601738824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,10240,0.46804534064398867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,10240,0.1529715591006809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,10240,0.4111795425415039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,8192,0.427855994966295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,16384,0.587073802947998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,8192,0.3299164507124159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,12288,0.4417208830515544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,7168,0.26828355259365505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,8192,0.13548710611131456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,10240,0.36932534641689724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,7168,0.28853599230448407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,7168,0.1131039990319146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,6144,0.2592524422539605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,8192,0.3010968897077772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,6144,0.2634826766120063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,5120,0.19768710931142172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,6144,0.09540266460842556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,4096,0.1585999992158678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,7168,0.2635866536034478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,5120,0.20724799897935656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,5120,0.08077866501278348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,3584,0.164410670598348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,6144,0.23223021295335558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,4096,0.0683386656973097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,4096,0.1753244400024414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,5120,0.1926248868306478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,3072,0.1375617716047499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,3584,0.060031996832953564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,4096,0.1580888827641805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,3584,0.1593831115298801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,2560,0.12432266606224908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,3072,0.12771466043260363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,3072,0.053433779213163585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,3584,0.14118844932980007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,2560,0.047255999512142606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,2048,0.08509422010845608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,2560,0.11018221908145481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,1536,0.062168002128601074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,3072,0.1225102212693956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,2048,0.04048000110520257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,2048,0.09125777747895981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,2560,0.10526044501198663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,1024,0.051098667913013034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,1536,0.033478223615222506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,1536,0.07242221964730157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,2048,0.08796444204118516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,768,0.041308446062935725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,1024,0.026884444885783728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,1536,0.07231911023457845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,512,0.027335110637876723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,1024,0.055663108825683594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,768,0.023767110374238756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,256,0.018548443913459778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,768,0.045280890332327954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,1024,0.056271996763017446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,512,0.021497777766651575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,128,0.017285333739386666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,512,0.03885866536034478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,768,0.04758755697144402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,256,0.019503111640612285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,64,0.01441333360142178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,512,0.04081688986884223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,256,0.03427377674314711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,32,0.014531556102964612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,128,0.03195555673705207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,128,0.019190222024917603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,256,0.03307466705640157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,128,0.025624889466497634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,64,0.03167288833194309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,51200,2.01688363817003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,32,0.03154755632082621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,51200,1.693894174363878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,65536,0.75018310546875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,65536,2.1951306660970054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,51200,1.689077377319336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,65536,2.124991946750217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,16384,0.48435020446777344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,12288,0.3713200092315674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,16384,0.18759644031524658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,16384,0.7450790935092502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,51200,0.5742479960123698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,12288,0.1418257819281684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,12288,0.3953813446892633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,10240,0.3474888801574707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,10240,0.12046666940053304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,10240,0.3811199929979112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,8192,0.30248623424106175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,12288,0.35041244824727374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,8192,0.25758666462368435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,8192,0.1017200019624498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,16384,0.4597902297973633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,10240,0.29646844334072536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,7168,0.23418577512105307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,7168,0.08610488971074422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,8192,0.23574577437506783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,6144,0.07564622163772583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,7168,0.2092142237557305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,6144,0.20252266195085314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,5120,0.18142222033606636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,6144,0.18126132753160265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,5120,0.16344532701704237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,5120,0.06461866696675618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,6144,0.19321866830190024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,4096,0.12399289343092178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,7168,0.22466667493184408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,5120,0.15363377994961208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,3584,0.11275822586483425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,4096,0.13407822450002035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,51200,1.5423466364542644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,4096,0.053973333703147046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,3584,0.1160480048921373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,3072,0.11264267232682969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,3584,0.048162665632036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,2560,0.09962133566538493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,3072,0.10142755508422852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,3072,0.049603554937574595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,3584,0.11209600501590306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,2048,0.0710444450378418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,2560,0.04089955488840739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,2560,0.08825333250893487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,2048,0.07323555813895331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,2048,0.033013333876927696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,2560,0.08350577619340684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,1536,0.0558204452196757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,2048,0.07244266404045953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,1536,0.02716355522473653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,1024,0.04440977838304308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,1536,0.058456891112857394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,1536,0.0593128866619534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,1024,0.020398222737842135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,4096,0.12668177816602919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,1024,0.045369777414533824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,768,0.028204444381925795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,512,0.022665777140193518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,768,0.01854311095343696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,1024,0.047480000389946826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,768,0.03642222285270691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,256,0.018760000665982563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,512,0.030415111117892798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,768,0.04074044360054864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,128,0.01567555632856157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,512,0.03473511007097032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,256,0.026875555515289307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,65536,2.0456506941053605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,64,0.014300444059901767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,256,0.028014222780863445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,128,0.015544889701737298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,32,0.014899555179807874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,128,0.02574666672282749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,64,0.025306666890780132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,128,0.02128977742460039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,32,0.025220443805058796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,65536,0.6036328739590114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,51200,0.4571777979532878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,65536,0.2559431129031711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,65536,0.6895635392930773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,256,0.015626667274369132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,16384,0.1492862171596951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,16384,0.17154933346642387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,51200,0.20162222120496961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,51200,0.5523529052734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,16384,0.06667377551396687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,16384,0.16520355807410345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,12288,0.12866577837202284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,512,0.016795555750528973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,12288,0.05172888769043816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,10240,0.1038657757971022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,12288,0.12740000089009604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,10240,0.0447306666109297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,10240,0.10800177521175808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,8192,0.0813973347345988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,65536,0.6532613436381022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,7168,0.08633510933982001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,10240,0.10907022158304851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,51200,0.5032871034410265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,8192,0.03658311234580146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,8192,0.0877386662695143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,6144,0.06173066960440742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,7168,0.035550223456488714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,6144,0.06821066803402372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,8192,0.08909689055548774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,6144,0.029873778422673542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,7168,0.08233155806859334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,5120,0.051226665576299034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,7168,0.08093955781724718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,4096,0.042400889926486544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,6144,0.07192355394363403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,5120,0.025211556090248957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,5120,0.058061334821912974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,3584,0.044565333260430225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,4096,0.02158222264713711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,4096,0.04851288927925957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,3072,0.037471999724706016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,5120,0.06328711244795057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,3584,0.019545776976479422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,3584,0.04276444514592489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,4096,0.05291644401020474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,2560,0.03160355488459269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,3072,0.018016000588734944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,3584,0.047036445803112455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,3072,0.037795556916130915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,2048,0.025789333714379206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,2560,0.016355555918481614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,2560,0.032518221272362605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,2560,0.03793777690993415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,1536,0.01998488936159346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,2048,0.027635556128289964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,3072,0.09790400001737808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,2048,0.03263555632697211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,1536,0.012753778033786349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,1024,0.014549333188268872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,1536,0.02277244461907281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,1536,0.028241776757770117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,1024,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,768,0.011919111013412476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,1024,0.018276444739765592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,768,0.009907555249002244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,512,0.0092257774538464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,768,0.015499555402331881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,12288,0.1372577746709188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,768,0.020601777566803824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,3072,0.04366666740841336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,512,0.009336888790130615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,256,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,512,0.012774222426944308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,128,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,2048,0.01460000044769711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,512,0.017495110630989075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,256,0.011767110890812345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,64,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,128,0.008729777402347988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,256,0.01516177753607432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,32,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,128,0.01239377756913503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,32,0.011338666909270816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,1024,0.023066666391160753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,65536,0.45583465364244247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,51200,0.3670328987969293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,65536,0.5209751129150391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,256,0.00869511150651508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,65536,0.2239217758178711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,16384,0.13912533389197454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,128,0.011391110718250275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,51200,0.4212951130337185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,51200,0.16763199700249565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,16384,0.1416977776421441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,12288,0.08594222201241387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,16384,0.05338133374849955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,10240,0.10379022359848022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,12288,0.04358399907747904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,16384,0.12946311632792154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,12288,0.10185688734054565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,8192,0.07286844650904338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,10240,0.03862488932079739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,10240,0.08488177590899998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,12288,0.10137155320909287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,7168,0.07164533270729913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,65536,0.5059502389695909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,8192,0.028573334217071533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,8192,0.06943466928270128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,10240,0.08688355816735162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,6144,0.05464444557825724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,7168,0.025552888711293537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,7168,0.06230755647023519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,8192,0.07257866859436035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,5120,0.05116622315512764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,51200,0.3901200029585097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,6144,0.02304711110062069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,7168,0.06633066468768649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,64,0.011404444773991903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,4096,0.04268088936805725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,5120,0.021868444151348535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,3584,0.036208000448015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,5120,0.04717955655521817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,4096,0.017909333109855652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,6144,0.058040002981821694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,4096,0.039317333035998873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,5120,0.04964355627695719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,3072,0.0315688881609175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,3584,0.01642933322323693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,3584,0.035763555102878146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,2560,0.02680622206793891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,4096,0.043362667163213096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,3072,0.015226667126019796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,2048,0.021985777550273474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,3072,0.031152000029881794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,3584,0.03943822118971083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,2560,0.014144889182514615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,1536,0.015295111470752291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,3072,0.035192890299691096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,2560,0.026820444398456152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,2048,0.022612444228596155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,2560,0.031424889961878456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,1024,0.012378666963842181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,1536,0.010666666759385003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,1536,0.018925334016482036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,2048,0.02753511071205139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,768,0.010772444307804108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,1024,0.00906755526860555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,1024,0.015298666225539314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,1536,0.023706666297382776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,512,0.008615111311276754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,768,0.013034666577974955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,1024,0.01993600030740102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,768,0.01719199948840671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,512,0.008282666405042013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,512,0.011014222270912595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,256,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,256,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,128,0.005821333163314396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,256,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,2048,0.012688889271683164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,256,0.014078223043017916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,64,0.005396444350481033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,128,0.009711110757456886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,32,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,128,0.01126488877667321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,6144,0.05494133300251431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,64,0.009725333087974125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,32,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,768,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,65536,0.4290142324235704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,51200,0.3276924557156033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,65536,0.4511600070529514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,512,0.01533600025706821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,16384,0.12180799908108181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,128,0.00811644477976693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,51200,0.14067910777197942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,16384,0.11287466684977214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,12288,0.0926284458902147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,16384,0.0488373339176178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,65536,0.4895048671298557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,12288,0.03825689024395413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,16384,0.1276364458931817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,12288,0.08616799778408474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,10240,0.07969688706927829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,8192,0.05600621965196398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,65536,0.17790488402048746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,51200,0.3783288796742757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,10240,0.03313510947757297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,10240,0.07235111130608453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,12288,0.09958222177293564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,7168,0.058520886633131236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,8192,0.027149332894219294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,8192,0.0596382220586141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,10240,0.08575733502705891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,6144,0.048807998498280845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,7168,0.024703999360402424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,51200,0.3581582175360785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,5120,0.04128088884883457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,7168,0.052925331725014575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,8192,0.07173689206441243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,6144,0.022082666556040447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,4096,0.028246220615175035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,6144,0.04685511191685995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,7168,0.06330221891403198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,5120,0.019648889700571697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,5120,0.04010577665434943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,4096,0.01721244388156467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,4096,0.033194667763180204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,3072,0.024990222520298425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,3584,0.0296053323480818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,3584,0.016006223029560514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,4096,0.041914665036731295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,2560,0.019638222124841478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,3072,0.014838222000333997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,3584,0.03839377893341912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,2560,0.013530666629473368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,2560,0.023231110639042322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,2048,0.016812443733215332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,3072,0.03436355458365546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,2560,0.03036710951063368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,2048,0.011810666984981961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,1536,0.013099555340078143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,2048,0.01993688941001892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,2048,0.026828444666332666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,3584,0.02439466615517934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,1024,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,1536,0.016768000192112394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,6144,0.056780444251166455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,768,0.008389333056079017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,5120,0.05001777741644117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,1536,0.02323377794689602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,1024,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,1024,0.013422222601042854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,512,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,768,0.00850311087237464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,768,0.011401777466138205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,3072,0.026563555002212524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,1024,0.019550222489568923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,512,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,512,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,768,0.017080000705189176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,128,0.004527111020353106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,256,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,512,0.014472888575659858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,256,0.009329777624871995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,64,0.004508444418509801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,32,0.00452888881166776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,128,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,128,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,256,0.013407111167907715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,128,0.010642666783597736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,64,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,1536,0.010423999693658618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,32,0.008980444735950893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,65536,0.25489510430230033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,51200,0.2004826731152005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,65536,0.13551644484202066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,65536,0.36301777097913956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,16384,0.09539022048314412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,256,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,51200,0.29514577653672963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,12288,0.07161066929499309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,51200,0.10730489095052083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,16384,0.09205866522259182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,16384,0.038033776813083224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,10240,0.05977777640024821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,12288,0.02956977817747328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,12288,0.07038044267230563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,16384,0.09318488836288452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,8192,0.04755999975734287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,51200,0.26912887891133624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,10240,0.06010933054818047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,65536,0.3452746603224013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,10240,0.025360888904995386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,12288,0.07531555493672688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,7168,0.04253155655331082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,8192,0.021527111530303955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,8192,0.049039999643961586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,7168,0.01957244508796268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,10240,0.06437600321239896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,7168,0.04373777906099955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,5120,0.025179555018742878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,8192,0.054099556472566396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,6144,0.017829333742459614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,4096,0.02514044443766276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,7168,0.04924711253907946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,5120,0.015957333975368075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,5120,0.03319999906751845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,3584,0.02208799951606327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,6144,0.04318933354483711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,4096,0.014455111490355598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,4096,0.02789777848455641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,5120,0.03847022189034356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,3072,0.01925244430700938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,3584,0.013377777404255338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,4096,0.03300177719857957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,2560,0.015186667442321777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,3072,0.01239733315176434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,3584,0.030214223596784804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,3072,0.022450667288568284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,2048,0.013978666729397245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,3072,0.027450667487250432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,2560,0.019697777099079557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,2560,0.01124888906876246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,6144,0.02995377779006958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,1536,0.01109866632355584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,2048,0.00981866651111179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,2048,0.01696622206105126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,1024,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,2560,0.02509422269132402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,1536,0.014461333552996317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,1536,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,768,0.007354666789372762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,2048,0.022223111655977037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,1536,0.019588443968031142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,1024,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,1024,0.011371555427710215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,512,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,768,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,1024,0.01648622254530589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,256,0.004535111288229625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,3584,0.025053333905008104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,512,0.008799110849698385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,512,0.007351111206743453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,128,0.003929777691761653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,512,0.01311377767059538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,256,0.00865955568022198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,128,0.008301333420806462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,128,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,256,0.01204533295498954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,32,0.003839111162556542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,128,0.009973333113723332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,6144,0.03882844580544366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,64,0.008303110798199972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,32,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,65536,0.30242400699191624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,768,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,65536,0.12962578402625188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,51200,0.2377164363861084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,65536,0.3410880035824246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,768,0.01477777792347802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,256,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,16384,0.08351377646128337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,16384,0.0872924460305108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,51200,0.2675164540608724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,16384,0.0370782216389974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,12288,0.06823911269505818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,51200,0.10252710845735337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,65536,0.33861600028143984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,16384,0.09371822410159641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,10240,0.053039110369152494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,12288,0.06722222434149848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,12288,0.02886577778392368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,51200,0.2654462125566271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,8192,0.043939554029040866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,10240,0.02517333295610216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,12288,0.07396533091862996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,10240,0.057258665561676025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,8192,0.047055999437967934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,7168,0.0389395554860433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,8192,0.02107644412252638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,10240,0.06341333521736993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,6144,0.033810665210088096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,7168,0.019415999452273052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,8192,0.053073777092827686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,7168,0.041999111572901406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,5120,0.02862311071819729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,6144,0.017642666896184284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,6144,0.03712266683578491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,64,0.0035217776894569397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,4096,0.023940443992614746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,5120,0.03196622265709771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,5120,0.015928000211715698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,3584,0.02167466613981459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,4096,0.014091556270917257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,5120,0.03784622086419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,4096,0.026978666583697002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,3072,0.019343111250135634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,3584,0.013165333204799227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,3584,0.024599111742443506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,4096,0.032637334532207914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,2560,0.016739555531077914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,3072,0.02180888917711046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,3072,0.012082666986518435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,3584,0.0295688874191708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,2048,0.014349333114094205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,2560,0.019321777754359774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,3072,0.027099554737408955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,2048,0.016501333978441026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,2048,0.009354666703277165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,2560,0.02457155618402693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,1536,0.011857777833938599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,7168,0.04758844441837735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,1024,0.009381333159075843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,6144,0.043439110120137535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,2048,0.022044445077578228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,1536,0.013612444202105204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,768,0.00831288927131229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,1024,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,1024,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,1536,0.018927110566033255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,768,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,1024,0.015927111109097798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,768,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,256,0.004836444639497333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,512,0.008642666869693333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,768,0.014533332652515836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,2560,0.010972444381978778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,256,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,128,0.004232888834344016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,256,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,512,0.013057777451144325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,64,0.0038942222793896994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,128,0.0069013333155049225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,256,0.012325333224402534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,32,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,128,0.007720000214046902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,1536,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,64,0.007727999654081132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,128,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,32,0.0076657773719893555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,512,0.006355555521117316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,51200,0.21011911498175728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,65536,0.12219466103447808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,65536,0.29849423302544487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,16384,0.0779235561688741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,51200,0.23125688234965006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,512,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,51200,0.09695733255810207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,12288,0.050400889582104154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,16384,0.07795200082990858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,16384,0.03530755639076233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,10240,0.04548089040650261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,12288,0.02795999911096361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,12288,0.06014933188756307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,65536,0.33707377645704484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,8192,0.03947644432385763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,51200,0.2651217778523763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,10240,0.024549333585633173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,10240,0.05094666613472832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,12288,0.07403733332951863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,7168,0.03256977929009332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,65536,0.2504968908098009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,8192,0.0206657780541314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,8192,0.04174133472972446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,6144,0.03065511253145006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,10240,0.06338133414586385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,7168,0.018903111418088276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,8192,0.052890668312708534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,5120,0.023362666368484497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,6144,0.03314133485158285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,6144,0.017258667283587985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,7168,0.04754755563206143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,4096,0.02051555613676707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,6144,0.04275022281540764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,5120,0.015607110328144498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,5120,0.02884355518552992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,3584,0.01811644434928894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,4096,0.01348977784315745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,5120,0.03724799884690179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,16384,0.09256711271074082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,3072,0.01646399994691213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,3584,0.012642666697502136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,4096,0.03211911188231574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,3072,0.019888000355826486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,2560,0.014296889305114746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,3584,0.029264890485339697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,3072,0.011505777637163797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,7168,0.037434667348861694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,3072,0.027070222629441157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,2048,0.012133333418104382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,2560,0.017487110363112558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,2560,0.024002666274706524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,2048,0.008979555633332994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,1536,0.009368889033794403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,2048,0.015088889333936902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,1024,0.007804444266690149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,2048,0.0218577782313029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,1536,0.008543111383914948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,4096,0.02432800001568264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,1536,0.012457778056462606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,768,0.006710222197903528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,1024,0.007896888587209914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,1024,0.01032977799574534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,512,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,1536,0.018221333622932434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,3584,0.022343110707071092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,768,0.007300444775157505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,768,0.009201777478059133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,256,0.004570666700601578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,1024,0.01551466683546702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,512,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,128,0.003903111235962974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,512,0.008046222229798635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,768,0.014683556225564746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,256,0.006945778098371293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,256,0.0075911109646161394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,64,0.0038159998754660287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,2560,0.010593777729405297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,512,0.013003555436929068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,32,0.0038817777401871155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,128,0.006691555596060223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,128,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,256,0.012191111014948951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,65536,0.1754524442884657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,64,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,32,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,51200,0.16098754935794407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,16384,0.05840888950559828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,65536,0.11939911047617595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,65536,0.2559893396165636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,51200,0.09534222549862331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,51200,0.19849599732293022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,12288,0.04662222332424588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,16384,0.0339955555068122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,16384,0.06695644723044501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,10240,0.03874489002757602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,12288,0.027038221557935078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,65536,0.33991021580166286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,16384,0.091903997792138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,8192,0.026872000760502283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,10240,0.02353333267900679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,10240,0.043954667117860585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,12288,0.07288889090220134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,7168,0.02716800073782603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,8192,0.01992177797688378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,8192,0.03650488787227207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,10240,0.06217066446940104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,128,0.009682666924264696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,6144,0.022435555855433147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,7168,0.03292266527811686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,7168,0.018311111463440787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,5120,0.0199297782447603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,8192,0.05203466614087423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,6144,0.01686844395266639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,6144,0.02931555443339878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,7168,0.046646220816506274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,4096,0.01646399994691213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,5120,0.01482222146458096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,5120,0.02566844390498267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,6144,0.041598220666249595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,3584,0.014645333091417948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,4096,0.013129777378506131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,5120,0.03678222166167365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,4096,0.021740444832377966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,12288,0.05149155524041918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,3072,0.012915555801656512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,3584,0.01963200006220076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,3584,0.011741333537631564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,4096,0.031138665146297876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,2560,0.011284444067213269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,3584,0.02843466732237074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,3072,0.011384888655609555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,2048,0.009320889082219865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,3072,0.026174222429593403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,2560,0.015478221906556023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,2560,0.023354666100607976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,2048,0.013447110851605734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,1536,0.008264888491895463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,2048,0.020294222566816542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,1536,0.011156444748242697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,1024,0.006048888795905643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,1536,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,1536,0.017907554904619854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,51200,0.2636231051550971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,768,0.005382222019963794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,1024,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,1024,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,3072,0.017878222796652053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,512,0.00433688900536961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,768,0.008343111309740279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,768,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,1024,0.01512888901763492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,256,0.003830222205983268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,2560,0.009983110758993361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,512,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,768,0.014217777384652032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,512,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,2048,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,256,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,256,0.006682666639486949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,64,0.0031928889867332247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,512,0.012740444805887012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,32,0.0032293332947625052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,128,0.006700444552633498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,256,0.011764444410800934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,64,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,65536,0.17432799604203966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,32,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,51200,0.1612515581978692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,65536,0.21290577782524955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,65536,0.07604800330268012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,16384,0.052390221092436046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,51200,0.060702224572499595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,51200,0.16830933094024658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,65536,0.19605778323279488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,128,0.0033697777738173804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,16384,0.05647644731733534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,16384,0.0225608895222346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,51200,0.15560089217291936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,10240,0.0329786671532525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,12288,0.043741332160102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,16384,0.05763555235332913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,128,0.009675555759006077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,8192,0.026399999856948853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,10240,0.016472889317406546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,12288,0.04673422376314799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,10240,0.037576887342664934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,7168,0.023678221636348303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,8192,0.014204444156752693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,10240,0.041161778900358416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,8192,0.03118577930662367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,6144,0.02013422217633989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,7168,0.028205331828859117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,7168,0.013439999686347114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,8192,0.034781333472993635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,5120,0.016505777835845947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,6144,0.01237422227859497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,12288,0.03579377796914842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,6144,0.02535733415020837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,7168,0.03154844376775954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,4096,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,5120,0.02222577730814616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,5120,0.011057777537239922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,3584,0.013080000049538083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,6144,0.028764443265067205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,128,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,4096,0.019153777096006606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,4096,0.009736888938479954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,3072,0.011407111254003314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,5120,0.025842666625976562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,3584,0.009229333036475712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,2560,0.010132444401582083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,3584,0.017320000463061862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,3072,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,3072,0.015420445137553744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,3584,0.02107911143038008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,2048,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,2560,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,3072,0.018955555227067735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,1536,0.006680000159475539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,2048,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,2048,0.011700444751315646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,2560,0.017276444368892245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,1024,0.0052968888647026485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,2048,0.015808888607554965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,1536,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,1024,0.00869422240389718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,1536,0.014504000544548035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,768,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,1024,0.006798222247097228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,1024,0.012692444854312472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,4096,0.022636445032225713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,768,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,768,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,768,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,2560,0.013439111411571503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,512,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,256,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,512,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,512,0.010967111421955956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,128,0.0028773333049482773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,256,0.006106666806671355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,256,0.00665511108107037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,256,0.011058666639857821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,128,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,128,0.0063031112982167145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,64,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,32,0.002867555452717675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,12288,0.018576888574494254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,128,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,65536,0.1480977800157335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,64,0.0063795554969045855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,32,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,51200,0.11663111050923665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,512,0.0038773334688610504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,16384,0.042018665207756885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,65536,0.07253599829143949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,65536,0.20904178089565703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,51200,0.058386663595835366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,51200,0.16330577267540827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,16384,0.02199199961291419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,16384,0.055092444022496544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,51200,0.15255378352271184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,10240,0.029417776399188574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,1536,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,12288,0.018201778332392376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,10240,0.016410667035314772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,12288,0.04516266783078512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,8192,0.023961777488390606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,10240,0.036509331729676985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,10240,0.04016266597641839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,8192,0.01408622165520986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,8192,0.030301332473754883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,7168,0.020619556307792664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,8192,0.03470755616823832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,7168,0.01312622262371911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,6144,0.018008000320858426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,7168,0.02755022214518653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,12288,0.03659644391801622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,6144,0.011924444801277585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,7168,0.03120533294147915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,65536,0.1904888947804769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,6144,0.024829333027203877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,4096,0.012959111067983838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,16384,0.057942224873436816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,5120,0.010696888797812991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,6144,0.028424888849258423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,3584,0.011392888923486074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,5120,0.021664889322386846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,4096,0.01864888932969835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,4096,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,5120,0.0251226673523585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,3072,0.010683555569913654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,3584,0.009023110899660323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,3584,0.016710221767425537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,2560,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,4096,0.02186311119132572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,3584,0.020669332808918424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,3072,0.008579555484983657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,3072,0.014834667245546976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,2048,0.008276444342401292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,3072,0.018710222509172227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,2560,0.01309511148267322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,1536,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,2048,0.011327111058764987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,2048,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,5120,0.015609777635998197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,1024,0.005530666559934616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,2048,0.015494222442309061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,1536,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,1536,0.009723555710580613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,768,0.004809777769777509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,12288,0.04258666767014397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,1024,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,1536,0.013757333159446716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,1024,0.00833511104186376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,512,0.004187555362780889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,768,0.006572444405820634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,768,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,256,0.0034959999223550162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,1024,0.012461333639091916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,512,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,512,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,128,0.0031884445084465873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,256,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,512,0.010692444112565784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,2560,0.008221333225568136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,256,0.010325333310498131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,64,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,128,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,2560,0.01719111038578881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,32,0.0029031110720502008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,128,0.008632889224423302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,64,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,65536,0.15316088994344076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,32,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,51200,0.10921511385175918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,65536,0.07060266865624322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,65536,0.17134666442871094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,768,0.01178311142656538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,51200,0.05722666449016995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,51200,0.1355031066470676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,65536,0.19149422645568848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,256,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,12288,0.03142755561404758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,128,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,16384,0.021664889322386846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,51200,0.15162489149305555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,16384,0.04577689038382637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,10240,0.026968000663651362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,12288,0.018011555075645447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,12288,0.03559111224280463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,8192,0.020616888999938965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,10240,0.03089155422316657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,10240,0.016196444630622864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,12288,0.045309331681993276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,7168,0.018602665927675035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,8192,0.013938667045699226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,8192,0.026035555534892615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,7168,0.012493333054913415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,8192,0.034182220697402954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,6144,0.016505777835845947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,7168,0.023349333140585158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,16384,0.04344177908367581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,5120,0.014192000031471252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,6144,0.011352889239788055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,7168,0.030939555830425684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,6144,0.021294222937689886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,4096,0.011930666863918304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,5120,0.010670222342014313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,5120,0.01868355605337355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,6144,0.02810844447877672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,3584,0.010749333434634738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,16384,0.057188444667392306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,4096,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,5120,0.02552977701028188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,3072,0.009826666778988307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,3584,0.014473777678277759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,3584,0.008768888811270395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,2560,0.008711111214425828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,4096,0.022117333279715642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,10240,0.04016622238688999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,3072,0.013066666821638743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,3072,0.008348444269763099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,2048,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,2560,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,3072,0.018428444862365723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,1536,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,2048,0.009993777506881291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,2048,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,2560,0.016827555166350473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,1024,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,1536,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,2048,0.015156444576051501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,1536,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,768,0.004555555681387584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,1024,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,1024,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,1536,0.013746666411558786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,512,0.003847111016511917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,768,0.0069884442620807225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,4096,0.015620445211728414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,768,0.006522666662931442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,1024,0.012422222230169507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,256,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,512,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,512,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,768,0.011685333318180509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,128,0.002883555574549569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,3584,0.020000000794728596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,256,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,256,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,64,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,512,0.01072533345884747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,128,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,128,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,2560,0.011198221809334226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,65536,0.12433421611785889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,128,0.008285333712895712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,32,0.006325333482689328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,51200,0.1180444426006741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,65536,0.07015289200676812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,51200,0.1292062203089396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,65536,0.19115377797020805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,16384,0.03849155704180399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,51200,0.15038577715555826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,16384,0.04442222250832451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,32,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,16384,0.056836446126302086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,12288,0.02812444501452976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,256,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,64,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,12288,0.017621333400408428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,10240,0.024051555328898962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,12288,0.03491466575198703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,65536,0.16522578398386636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,12288,0.045062220758861966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,10240,0.015876443849669564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,8192,0.019056888090239633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,51200,0.05644622113969591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,8192,0.013090666797426013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,7168,0.017351999878883362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,8192,0.025443555580245122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,8192,0.03366577625274658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,7168,0.012144000165992312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,7168,0.023046221998002794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,6144,0.015099555253982544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,6144,0.011414222419261932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,5120,0.013219555219014486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,6144,0.02083200050724877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,6144,0.027649778458807204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,4096,0.011339555184046427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,5120,0.010001777774757808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,10240,0.030294222964180842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,5120,0.02539199921819899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,3584,0.010205333431561789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,4096,0.015088889333936902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,10240,0.03897510965665182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,4096,0.021486222743988037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,3584,0.00869066682126787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,3584,0.013774221969975365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,3584,0.01984444426165687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,3072,0.01238844460911221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,16384,0.02102311121092902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,2560,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,3072,0.008296888735559251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,3072,0.018410666121376883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,2048,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,2560,0.00793511089351442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,2560,0.011053333679835001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,5120,0.018264888061417472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,1536,0.006124444305896759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,2048,0.007445333732499017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,2560,0.016540444559521146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,2048,0.009687999884287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,1024,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,1536,0.0069857777820693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,2048,0.015159999330838522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,1536,0.008709333009190029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,768,0.0038862224254343244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,1024,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,1024,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,1536,0.013800000150998434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,512,0.003510222252872255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,768,0.006423111177153057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,768,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,1024,0.012675555215941535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,7168,0.03074666526582506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,512,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,768,0.011767110890812345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,512,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,128,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,256,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,256,0.005967111223273807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,128,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,128,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,256,0.010051555931568146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,64,0.002505777817633417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,32,0.002537777854336633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,4096,0.009154666629102495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,128,0.008426666259765625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,64,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,32,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,3072,0.009135111338562435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,51200,0.07429689168930054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,65536,0.0682355562845866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,65536,0.1284524467256334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,16384,0.033181332879596286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,256,0.0029022221763928733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,51200,0.10172888967725967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,51200,0.05519377854135302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,65536,0.18089689148796928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,12288,0.0258879992696974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,16384,0.02088533341884613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,16384,0.034860445393456355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,512,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,10240,0.021474666065639917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,12288,0.01732711162832048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,12288,0.027658666173617046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,16384,0.05371377865473429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,8192,0.01774577796459198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,10240,0.015468445089128284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,12288,0.04288888971010844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,10240,0.024267555938826665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,7168,0.015743111570676167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,65536,0.11159643861982559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,8192,0.012669333153300814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,8192,0.020642666353119742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,10240,0.037422223223580256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,6144,0.013967111706733704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,7168,0.018889778190188937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,7168,0.011706666813956367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,8192,0.031967110104031034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,5120,0.012327111429638334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,6144,0.01074044406414032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,7168,0.029475556479560003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,4096,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,5120,0.0148053334818946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,5120,0.009986666341622671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,6144,0.02667288813326094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,3584,0.009395555489593083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,5120,0.023665777511066858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,4096,0.00906044410334693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,4096,0.01278133359220293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,51200,0.14486311541663274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,3072,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,3584,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,3584,0.011767999993430244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,2560,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,4096,0.020071999894248117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,3072,0.008328888979223039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,3072,0.010421333213647207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,2048,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,3584,0.01864088906182183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,3072,0.017296888762050204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,2560,0.009322666459613377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,1536,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,2048,0.007605333295133378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,2048,0.008369777765538957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,6144,0.016890666551060148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,2048,0.01443288889196184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,1536,0.006983111302057902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,1024,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,1024,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,1536,0.013407111167907715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,768,0.0039057777159743835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,1024,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,1024,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,768,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,512,0.0032248888164758682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,768,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,768,0.011348444554540845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,512,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,512,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,256,0.003176888864901331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,2560,0.007921777665615082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,128,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,512,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,256,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,2560,0.016028443972269695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,256,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,64,0.0026275554878844153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,32,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,128,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,1536,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,256,0.00998577806684706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,65536,0.08052533202701144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,64,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,128,0.007686222592989604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,32,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,51200,0.06450311342875163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,16384,0.02606311109330919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,65536,0.12175110975901286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,65536,0.06721422407362196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,51200,0.053598221805360585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,12288,0.019326221611764696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,51200,0.09547822342978583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,16384,0.02027555637889438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,65536,0.17871556017133924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,16384,0.03414222266938951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,10240,0.01646222174167633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,12288,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,12288,0.027274666561020747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,16384,0.052784889936447144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,8192,0.013908444179428948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,10240,0.014112888111008538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,10240,0.023904000719388325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,7168,0.01236177815331353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,128,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,10240,0.036575112077924944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,8192,0.019786667492654588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,6144,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,7168,0.017889777819315594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,7168,0.011549333731333414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,8192,0.03170399864514669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,5120,0.00980888886584176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,6144,0.010515555739402771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,6144,0.016210666961140104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,7168,0.02886755598915948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,4096,0.008492444124486711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,5120,0.014532445205582513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,6144,0.025681777132882014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,4096,0.012366222010718452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,51200,0.14210577805836996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,5120,0.022893332772784766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,3584,0.007707555260923173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,4096,0.01977244516213735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,12288,0.042019555966059365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,3072,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,3584,0.011437333292431302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,3584,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,8192,0.012128888732857175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,2560,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,3584,0.018246221873495314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,3072,0.008216888540320927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,3072,0.01039644413524204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,2048,0.0053137776752312975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,2560,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,3072,0.01687733332316081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,2560,0.015473778049151102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,2048,0.0073520003093613525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,1536,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,2048,0.014476444986131458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,4096,0.009112888740168678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,1536,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,1536,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,1536,0.012877333495352002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,1024,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,768,0.003527111063400904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,1024,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,1024,0.012048000262843238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,768,0.006302222195598815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,2560,0.007803555164072249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,512,0.002891555635465516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,768,0.006312888943486744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,768,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,256,0.0026062221990691293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,5120,0.009770666559537252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,2048,0.008347555167145198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,512,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,128,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,256,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,256,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,1024,0.0038453332251972626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,64,0.002461333241727617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,128,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,128,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,256,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,65536,0.051182223690880664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,128,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,64,0.005518222020732031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,32,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,51200,0.04303555687268575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,65536,0.0811413327852885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,65536,0.06576533450020684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,16384,0.017838221457269456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,51200,0.05199466811286079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,51200,0.06388355626000299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,65536,0.17682488759358725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,512,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,12288,0.017614222235149808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,16384,0.023994666006830003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,51200,0.14049955209096274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,10240,0.015507555670208402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,512,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,12288,0.015829333000712924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,12288,0.01944266590807173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,16384,0.051983998881445996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,8192,0.013148444394270578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,10240,0.013855111267831592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,10240,0.016865778300497267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,12288,0.04141066802872552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,7168,0.01181422256761127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,8192,0.012084444363911947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,8192,0.01425333321094513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,6144,0.010597333312034607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,10240,0.03607733382119073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,7168,0.01312800000111262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,7168,0.011413333316644033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,8192,0.03120622369978163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,5120,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,6144,0.010538666612572141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,6144,0.012081777883900536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,4096,0.007920000288221572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,5120,0.011066666907734342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,5120,0.009649777577983009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,6144,0.025669333007600572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,3584,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,16384,0.019483556350072224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,5120,0.022431110342343647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,4096,0.008971555365456475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,4096,0.009843555589516958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,3072,0.006292444550328785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,3584,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,3584,0.009004444711738164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,2560,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,4096,0.01946311030122969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,3072,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,32,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,3072,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,2048,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,3584,0.01812533371978336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,2560,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,3072,0.016777777009540133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,2560,0.00777333312564426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,1536,0.0041759999261962045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,2048,0.007344889144102733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,2048,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,2560,0.015491555134455362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,1024,0.003552000141806073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,1536,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,1536,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,7168,0.027492443720499676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,768,0.0031715554909573663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,1024,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,1536,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,512,0.0028764444092909494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,768,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,768,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,1024,0.011415111521879831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,256,0.0025315555847353404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,768,0.010703110860453712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,512,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,512,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,128,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,256,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,256,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,512,0.010374222364690568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,64,0.002151111140847206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,128,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,32,0.0021848889688650766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,256,0.009704888694816166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,65536,0.04401155643992954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,64,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,128,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,32,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,2048,0.014095111025704278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,51200,0.03743911120626662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,65536,0.06411733229955037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,1024,0.006110222389300664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,65536,0.07931466897328694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,16384,0.015625778171751235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,51200,0.06252177556355794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,51200,0.050309333536359996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,12288,0.013772444592581855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,65536,0.1770284440782335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,16384,0.023358222511079576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,16384,0.018939556346999276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,51200,0.14020000563727483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,10240,0.012727999852763282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,16384,0.05128177669313219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,12288,0.01535377734237247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,12288,0.018931556079122756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,8192,0.012004444168673621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,128,0.005103111267089844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,10240,0.01407911048995124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,10240,0.016415110892719693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,12288,0.04119377666049533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,7168,0.011792889071835412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,8192,0.012067555553383298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,10240,0.03580177823702494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,6144,0.010424888796276517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,7168,0.011328000161382886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,5120,0.008621333373917473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,7168,0.013052444491121503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,8192,0.02994133366478814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,6144,0.010432889064153036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,4096,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,6144,0.011893333660231696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,7168,0.027436445156733196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,5120,0.009703111317422655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,3584,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,6144,0.02470933397610982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,4096,0.009009777671760982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,4096,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,5120,0.022172444396548804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,3072,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,3584,0.008328888979223039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,3584,0.009157333109113906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,4096,0.0188755558596717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,2560,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,3072,0.008247111406591203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,3072,0.008531555533409119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,3584,0.017845332622528076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,2048,0.004950222041871813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,2560,0.007671111159854465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,2560,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,3072,0.01685688893000285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,1536,0.0042239998777707415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,8192,0.014261333478821648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,2560,0.015463999576038785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,2048,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,1024,0.00351555562681622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,1536,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,1536,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,2048,0.014012444350454541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,768,0.003305777700410949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,1024,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,1024,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,5120,0.011016888750924004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,512,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,768,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,768,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,1024,0.011352889239788055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,256,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,512,0.006215111249023014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,768,0.011054221954610614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,512,0.005440000030729506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,128,0.002465777720014254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,256,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,256,0.005868444426192178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,512,0.010273777776294285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,64,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,256,0.009424000150627559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,128,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,768,32,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,128,0.005065777649482091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,65536,0.03203377789921231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,64,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,2048,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,128,0.007605333295133378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,32,0.005142222262091107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,65536,0.07801333400938246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,65536,0.0633173320028517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,16384,0.013436444103717804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,1536,0.013060444758998023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,51200,0.06184977955288357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,51200,0.048419555028279625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,65536,0.17703644434611002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,16384,0.018269333574506972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,16384,0.022631110416518316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,51200,0.1397671169704861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,10240,0.011730666789743634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,12288,0.018171555466122098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,12288,0.015228443675571017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,16384,0.05096266667048136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,8192,0.01241422196229299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,10240,0.013803555733627744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,10240,0.016182222300105624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,12288,0.0406968891620636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,7168,0.010775999890433418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,8192,0.01185066666867998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,10240,0.03492533498340183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,51200,0.025643555654419795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,6144,0.009992889232105678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,7168,0.011049778097205691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,7168,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,8192,0.02968177861637539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,5120,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,6144,0.01034400032626258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,7168,0.027056000298923914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,6144,0.011701333026091257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,12288,0.011136000355084738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,4096,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,5120,0.01069866700304879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,5120,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,3584,0.00657066661450598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,6144,0.02440888848569658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,4096,0.008714666797055138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,4096,0.009662222531106737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,3072,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,5120,0.021873777111371357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,3584,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,4096,0.018927110566033255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,2560,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,3072,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,3072,0.008269333177142674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,3584,0.017599110802014668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,2048,0.004536889079544279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,8192,0.01407911048995124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,2560,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,2560,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,1536,0.003913777983850903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,3072,0.01681866745154063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,2048,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,1024,0.0031964443624019623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,2048,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,1536,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,1536,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,2048,0.013935110635227628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,768,0.002946666752298673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,1024,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,1024,0.006010666489601135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,512,0.002569777684079276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,1536,0.012612444659074148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,768,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,768,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,256,0.002326222136616707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,512,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,512,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,768,0.010677333507272931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,128,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,3584,0.00904088881280687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,512,0.01035288886891471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,256,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,256,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,64,0.002114666625857353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,512,32,0.002159111201763153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,128,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,128,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,65536,0.023342221975326538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,256,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,64,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,32,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,2560,0.015173332558737861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,65536,0.05997777647442288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,65536,0.07710666788948907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,51200,0.06048711140950521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,51200,0.04843733376926846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,16384,0.01036888857682546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,65536,0.1766080061594645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,1024,0.011384000380833944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,12288,0.012077333198653327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,51200,0.13884354962242976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,16384,0.022104000051816303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,16384,0.04986044433381823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,12288,0.015311110350820752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,12288,0.017885333961910672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,10240,0.010934222075674268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,12288,0.039510223600599505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,10240,0.01382844481203291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,8192,0.009431111315886179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,10240,0.015840889679061044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,51200,0.018276444739765592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,7168,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,8192,0.01183644433816274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,8192,0.013784888717863293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,8192,0.02940888868437873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,7168,0.011047999891969891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,6144,0.008036444584528605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,16384,0.017978666557206046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,7168,0.026438222991095647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,5120,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,6144,0.011728000309732227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,6144,0.010370666782061258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,6144,0.024039111203617517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,4096,0.006672888994216919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,5120,0.00961511085430781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,5120,0.01038755559258991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,4096,0.00869066682126787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,5120,0.02164622147878011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,3584,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,128,0.00758755538198683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,4096,0.009354666703277165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,3072,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,10240,0.034437331888410784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,3584,0.008535111116038429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,4096,0.018684445155991446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,2560,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,3072,0.007931555310885111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,7168,0.012631110846996307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,2048,0.004576888763242298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,3584,0.01750577820671929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,2560,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,2560,0.007557333343558841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,1536,0.0038248888320393036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,3072,0.01644088824590047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,2048,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,2048,0.007019555403126611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,2560,0.015467555986510383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,1024,0.0032053333189752367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,1536,0.00685511115524504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,1536,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,2048,0.013542222479979197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,768,0.002814222127199173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,1024,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,1024,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,1536,0.012757333616415659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,512,0.00254844439526399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,768,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,1024,0.011119999819331698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,256,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,512,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,512,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,3584,0.008978666530715095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,128,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,256,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,256,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,3072,0.00832088871134652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,64,0.0018773333479960759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,128,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,256,32,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,128,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,256,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,65536,0.0165964447789722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,64,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,128,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,51200,0.0136817776494556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,32,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,65536,0.058335999647776283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,65536,0.07664444711473253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,16384,0.009256000320116678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,51200,0.04435999857054817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,51200,0.05805777841144138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,768,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,65536,0.176456888516744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,12288,0.008211555580298105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,16384,0.021793777743975323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,16384,0.017855111095640395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,51200,0.1387795607248942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,10240,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,12288,0.015117333994971381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,512,0.01034400032626258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,8192,0.007255111303594377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,16384,0.0493199990855323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,10240,0.013752000199423896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,10240,0.015853333804342482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,12288,0.039566222164365984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,7168,0.0069164443347189165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,10240,0.03393955694304572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,8192,0.013654222091039022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,8192,0.011782222323947482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,6144,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,8192,0.02915289004643758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,6144,0.011388444238238864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,7168,0.026495999760097925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,5120,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,6144,0.01038755559258991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,6144,0.02405244443151686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,768,0.010687111152542962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,4096,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,5120,0.010354667074150508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,5120,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,4096,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,5120,0.021300445000330608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,4096,0.009327111144860586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,3584,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,3584,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,4096,0.018234666850831773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,3584,0.008700444466537898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,7168,0.012423111332787408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,3584,0.01722400055991279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,2560,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,3072,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,3072,0.016159110599093966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,2560,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,2048,0.004254222330119875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,2560,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,12288,0.017881777551439073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,2560,0.0148053334818946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,1536,0.003863999827040566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,2048,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,1536,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,1536,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,1024,0.0031511110977994073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,2048,0.013467555244763693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,3072,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,768,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,7168,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,1024,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,3072,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,512,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,1024,0.011012444065676795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,768,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,512,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,256,0.0021439999755885866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,768,0.010802666346232096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,512,0.009648888475365108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,2048,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,256,0.005199111170238919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,128,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,256,0.009670221971140968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,128,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,1536,0.012859555582205454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,64,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,128,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,128,32,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,1024,0.006531555619504716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,65536,0.01482311056719886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,51200,0.012746666868527731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,128,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,16384,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,768,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,64,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,12288,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,32,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,65536,0.07463466458850436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,8192,0.005672889037264719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,51200,0.057597332530551486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,16384,0.02165688905451033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,512,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,12288,0.017665778597195942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,6144,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,10240,0.01571466690964169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,5120,0.00656444455186526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,4096,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,8192,0.013493333425786761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,3584,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,7168,0.012416000167528788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,3072,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,6144,0.011390222443474663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,256,0.005879111174080107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,2560,0.004971555537647671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,5120,0.0103502223889033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,2048,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,4096,0.009104889300134447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,1536,0.0035253332720862497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,1024,0.003061333257291052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,3584,0.008612444831265343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,768,0.0028577778074476454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,2560,0.007248000138335758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,512,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,2048,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,1536,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,256,0.002143111079931259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,1024,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,768,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,512,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,256,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,32,0.0017991111510329777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,7168,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,128,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,65536,0.027486221657858953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,51200,0.02219644520017836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,64,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,16384,0.008625778059164682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,32,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,12288,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,65536,0.07453511158625285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,10240,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,51200,0.05739733245637682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,16384,0.021663111117151048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,12288,0.01756355497572157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,7168,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,10240,0.015592000550693937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,6144,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,8192,0.013421333498424955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,128,0.0018053334206342697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,5120,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,4096,0.006029333505365584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,7168,0.01237155579858356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,3584,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,64,10240,0.006311999840868845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,6144,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,3072,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,2560,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,5120,0.01036622209681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,2048,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,3584,0.008656889200210571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,1536,0.003546666767862108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,3072,0.008011555506123437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,2560,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,2048,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,768,0.0028391112056043413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,1536,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,512,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,8192,0.00600000015563435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,1024,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,256,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,128,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,768,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,3072,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,64,0.0018053334206342697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,32,0.001812444378932317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,512,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,256,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,128,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,64,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,32,0.004954666727119022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,4096,0.00903555585278405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,16384,0.3968319892883301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,512,32,1024,0.0029146667155954572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,16384,0.4131608804066976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,12288,0.3253786563873291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,16384,0.22871200243631998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,12288,0.16182400120629206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,12288,0.3113831149207221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,10240,0.2760097715589735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,51200,1.3071635564168294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,10240,0.2717306613922119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,10240,0.13764622476365831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,8192,0.2300915453169081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,51200,1.3795217937893336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,7168,0.21964889102511934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,8192,0.20952267116970488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,6144,0.17072178257836235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,7168,0.18358666367001006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,51200,0.6770604451497396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,5120,0.14477245012919107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,6144,0.15840000576443142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,4096,0.13728533850775823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,5120,0.07193066676457723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,3584,0.12142844994862874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,5120,0.13340532779693604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,6144,0.08540799882676865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,4096,0.05928888585832384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,4096,0.10935644308725993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,3072,0.09540088971455891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,2560,0.0762266649140252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,3584,0.09539822075102065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,3584,0.05611111058129204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,3072,0.05201244354248047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,2048,0.06733955277336968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,3072,0.08304266797171699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,1536,0.05192977852291531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,2560,0.04048533240954081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,1024,0.03568177753024631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,2048,0.0619608892334832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,8192,0.11253510581122504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,1536,0.027884443600972492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,7168,0.09847288661532932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,1536,0.04745422138108147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,512,0.020596444606781006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,1024,0.022270222504933674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,1024,0.037031110790040754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,768,0.02017600006527371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,768,0.02996266550487942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,256,0.013021333350075616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,512,0.01850844422976176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,512,0.025397333833906386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,128,0.010317333042621613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,256,0.016813332835833233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,64,0.009197333620654212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,128,0.016619554824299283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,256,0.022171555293930903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,32,0.01202666676706738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,128,0.021486222743988037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,64,0.021317332983016968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,32,0.0210844443904029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,2560,0.071706665886773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,2048,0.03545777665244208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,768,0.028661333852344092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,16384,0.3383342160118951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,51200,1.0532586839463975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,65536,1.3465155495537653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,51200,1.0706969367133248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,12288,0.2533715565999349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,65536,0.7277226448059082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,51200,0.538432863023546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,12288,0.24370755089653862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,16384,0.16932977570427787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,10240,0.232044431898329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,12288,0.12946666611565485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,8192,0.1901279952791002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,10240,0.10825066434012519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,7168,0.15294044547610813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,65536,1.367438210381402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,10240,0.22126666704813638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,8192,0.1253137720955743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,8192,0.16421511438157824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,6144,0.12808177206251356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,16384,0.3333608839246962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,7168,0.14821599589453802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,7168,0.07723733451631334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,5120,0.1194266610675388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,6144,0.06713244650099012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,3584,0.08533689048555161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,5120,0.10530577765570746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,4096,0.04776800009939405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,4096,0.08761688735749985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,3584,0.0767493314213223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,3072,0.06858577993181017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,6144,0.12559821870591906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,3584,0.05090844300058153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,2560,0.06301866637335883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,2048,0.051069332493676074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,3072,0.03758222195837233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,2560,0.03348088926739163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,1536,0.03984444340070089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,3072,0.06649866369035509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,2048,0.029110223054885864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,2560,0.057465778456793896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,2048,0.04888355400827196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,1536,0.022904000348514978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,768,0.021839999490314062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,1024,0.027338667048348322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,1536,0.039176887936062284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,1024,0.018738667170206707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,1024,0.030585779084099665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,768,0.01696711116366916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,256,0.010882666541470421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,512,0.015842666228612263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,768,0.024549333585633173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,4096,0.09466489156087239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,128,0.008978666530715095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,512,0.020608888732062448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,256,0.014374223020341663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,32,0.007721777591440413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,256,0.018589332699775696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,128,0.014103111293580798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,128,0.017529777354664274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,5120,0.057688891887664795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,64,0.01754044493039449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,32,0.01756088932355245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,65536,0.3955830997890896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,51200,0.3110720051659478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,16384,0.11223555935753717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,65536,0.4373955461714003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,65536,0.23558134502834746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,51200,0.3511182202233209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,51200,0.19001600477430555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,12288,0.08534666564729478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,10240,0.07790222432878283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,12288,0.04830222328503927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,16384,0.06856533553865221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,12288,0.08418933550516765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,64,0.007368889119890001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,16384,0.1122702227698432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,8192,0.060686224036746554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,10240,0.07141244411468506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,7168,0.05235377616352505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,6144,0.04742044541570875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,8192,0.0340595543384552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,7168,0.030813333061006334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,8192,0.06069866816202799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,5120,0.040884445110956825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,7168,0.05231111248334249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,4096,0.03261244297027588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,6144,0.045660446087519325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,5120,0.02386933399571313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,3584,0.029687109920713637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,4096,0.02051999999417199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,5120,0.039529777235454984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,3072,0.02590755621592204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,4096,0.03276622295379639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,3584,0.018758222460746765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,512,0.01551555593808492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,3584,0.029345777299669053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,3072,0.017124444246292114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,2560,0.021593777669800654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,2048,0.017818666166729398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,3072,0.02612888813018799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,2560,0.015496888094478183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,2048,0.014064888159434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,2560,0.02273599969016181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,1024,0.010552888943089379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,10240,0.04095110959476895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,2048,0.01944177846113841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,1536,0.012106666962305704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,1536,0.016349333855840895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,1024,0.010103999740547603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,512,0.0075093333919843035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,1024,0.013306666579511432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,6144,0.02717688845263587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,768,0.009330666727489894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,256,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,768,0.011689778003427716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,128,0.004249777644872665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,512,0.009027555584907532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,256,0.008303110798199972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,512,0.009716444545321995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,64,0.0042408886882993905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,256,0.009336888790130615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,32,0.0042364444169733255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,128,0.008966222405433655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,64,0.008992888861232335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,32,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,65536,0.3045022222730848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,1536,0.014096889230940076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,51200,0.27655911445617676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,768,0.009012444151772393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,65536,0.337752898534139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,65536,0.18540710873074004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,16384,0.08575377861658733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,12288,0.06943200031916301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,51200,0.14390667279561362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,51200,0.2642951011657715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,16384,0.05256888932651944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,10240,0.05814311239454481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,16384,0.08797599871953328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,8192,0.044518222411473594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,12288,0.04288088944223192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,12288,0.07389599747127958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,7168,0.04103644357787238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,10240,0.03341155582004123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,128,0.008401778009202745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,8192,0.03220977716975742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,6144,0.03638044330808852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,7168,0.02472711106141408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,5120,0.02994133366478814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,8192,0.04707555638419258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,10240,0.05699200100368924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,4096,0.0243520008193122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,6144,0.022089777721299067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,3584,0.021151999632517498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,5120,0.019486222002241347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,6144,0.03715199894375271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,5120,0.031778666708204485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,3072,0.01909244391653273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,4096,0.017328000730938382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,2560,0.01590666671593984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,3584,0.018153778380817838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,2048,0.013436444103717804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,3072,0.014739554789331226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,3584,0.02388444377316369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,2560,0.013502221968438892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,3072,0.021514667405022517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,2560,0.018949333164427016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,2048,0.012071111136012606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,1536,0.010448888772063784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,2048,0.01633155511485206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,768,0.007928888830873702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,1536,0.013912000589900546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,1024,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,1024,0.011564444336626263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,512,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,768,0.008515555825498369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,256,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,768,0.009707555174827576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,512,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,128,0.0038942222793896994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,512,0.008640888664457533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,256,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,7168,0.044065776798460216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,64,0.003536000019974179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,256,0.008073777788215214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,32,0.0038204445607132385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,128,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,128,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,4096,0.02637333340115017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,64,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,32,0.00794666674402025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,65536,0.2803902096218533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,1024,0.009135111338562435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,51200,0.2358915540907118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,16384,0.08449421988593207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,65536,0.2964284420013428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,65536,0.1697902282079061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,51200,0.23340977562798393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,12288,0.057761775122748486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,16384,0.046225776274998985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,51200,0.13317067093319362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,10240,0.04849066668086582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,12288,0.03837777839766608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,8192,0.04068266683154636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,12288,0.0588959985309177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,10240,0.03119200136926439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,7168,0.0369031098153856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,10240,0.049867557154761426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,8192,0.02628799941804674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,6144,0.03331644336382548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,8192,0.041738665766186185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,7168,0.02372177773051792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,5120,0.02779910961786906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,7168,0.036836445331573486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,6144,0.02127733329931895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,4096,0.02260977857642704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,6144,0.03262133399645487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,5120,0.018852444158660043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,3584,0.020208888583713107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,4096,0.016703110602166917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,5120,0.027848889430363972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,3072,0.017761778500345018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,4096,0.023366222778956097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,2560,0.015377778146002026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,3584,0.015319110618697273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,2048,0.013163555827405719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,3072,0.014149333039919535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,2560,0.013443555268976422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,3072,0.01925244430700938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,1536,0.011122666299343109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,2560,0.01716711123784383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,2048,0.011295110815101199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,1024,0.009379555781682333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,1536,0.009690666364298927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,768,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,1536,0.012770666844315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,1024,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,512,0.006206222292449739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,1024,0.010236444572607676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,768,0.008357333640257517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,256,0.004520888957712385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,512,0.007838221887747446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,768,0.00907022257645925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,128,0.0038337777886125776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,512,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,256,0.007381333245171442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,1536,0.011170667078759937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,64,0.003506666670242945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,256,0.007582222421964009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,32,0.0038337777886125776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,128,0.0075395554304122925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,128,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,3584,0.021143111917707656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,65536,0.24292445182800293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,32,0.007355555891990662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,16384,0.07583555248048571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,51200,0.2104897763994005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,2048,0.015061333775520325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,65536,0.24383200539482963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,16384,0.06631910800933838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,12288,0.05043733451101515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,51200,0.10226400030983819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,16384,0.03587377733654446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,51200,0.19247821966807047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,10240,0.043980442815356784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,16384,0.06486666864818998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,8192,0.035164445638656616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,12288,0.028091556496090356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,7168,0.03197600113021003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,10240,0.02459377712673611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,8192,0.023678221636348303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,10240,0.04325244492954678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,6144,0.02738933265209198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,7168,0.019097778532240126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,5120,0.023415999280081853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,7168,0.03168977631462945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,6144,0.017370666066805523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,4096,0.01959466603067186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,6144,0.02814133299721612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,5120,0.015504888362354703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,64,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,5120,0.0240666667620341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,4096,0.014076444837782117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,3072,0.014827556080288358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,4096,0.02050933407412635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,3584,0.01311644415060679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,3584,0.018903111418088276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,2560,0.012400888734393649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,3072,0.012005333271291522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,3072,0.017063111066818237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,12288,0.04980977707438999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,2560,0.010687111152542962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,2560,0.01519555515713162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,1536,0.008740444150235917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,2048,0.00982844415638182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,8192,0.035585777627097234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,2048,0.013389333254761167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,1024,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,1536,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,768,0.005792888916201062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,1536,0.011238222320874533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,1024,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,512,0.004561777744028303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,3584,0.01682400041156345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,1024,0.009009777671760982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,768,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,768,0.008298666940795051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,512,0.007170666423108842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,128,0.003167111012670729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,512,0.007393777370452881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,256,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,256,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,64,0.0033440000067154565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,128,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,32,0.0033448889023727844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,128,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,2048,0.010807111031479306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,64,0.006899555524190267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,65536,0.23267645306057402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,51200,0.18250577979617646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,65536,0.12230843967861599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,16384,0.06284799840715197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,51200,0.16404622130923802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,256,0.003527111063400904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,51200,0.09656711419423421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,12288,0.04806844393412272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,65536,0.12464622656504314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,16384,0.055732442273033984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,10240,0.04046133491728041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,12288,0.02709688908523983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,12288,0.04309511184692383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,8192,0.033815999825795494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,10240,0.02392622166209751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,10240,0.03711466656790839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,7168,0.029829333225886028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,8192,0.020317332612143624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,6144,0.026017778449588354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,32,0.006816000160243776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,8192,0.03062044580777486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,7168,0.020741333564122517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,5120,0.021470222208235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,6144,0.017236444685194228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,4096,0.01811555524667104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,7168,0.027097778187857732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,3584,0.015830222103330824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,6144,0.02411199940575494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,5120,0.01548266741964552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,65536,0.20983466837141249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,3072,0.014298665854665967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,4096,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,5120,0.020970667401949566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,2560,0.012358222570684222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,3584,0.013260444833172692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,4096,0.018055111169815063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,2048,0.010644444160991244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,3584,0.016543111867374845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,2560,0.010573333336247338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,1536,0.008653333617581261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,3072,0.015097777048746744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,2560,0.013434666726324292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,16384,0.034712000025643244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,2048,0.012114666402339935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,768,0.005708444449636672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,1024,0.008018666671382057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,1024,0.008349333372380998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,512,0.004559111264016894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,768,0.00738844441043006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,768,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,256,0.004028444488843282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,512,0.007058666812049017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,512,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,128,0.0035413333939181436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,256,0.006687110910813014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,256,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,64,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,128,0.006700444552633498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,32,0.0031822222388452957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,128,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,64,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,2048,0.009391110804345872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,1024,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,65536,0.19155822859870064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,32,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,1536,0.00831288927131229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,1536,0.009718221922715506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,16384,0.05815822548336453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,65536,0.11568089326222737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,65536,0.19776444964938691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,51200,0.09229066636827256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,12288,0.04206577936808268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,51200,0.15609067016177708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,10240,0.03334133492575751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,16384,0.053545776340696544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,12288,0.026763555076387193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,12288,0.04163199994299147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,8192,0.029324445459577773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,10240,0.023229334089491103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,7168,0.025979555315441553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,8192,0.020215999748971727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,3072,0.01276000009642707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,8192,0.029733333322736952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,6144,0.022111111217074927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,7168,0.018326222896575928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,7168,0.026428444517983332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,5120,0.018801778554916382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,6144,0.016749334004190233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,4096,0.015306666493415833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,6144,0.02366133365366194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,5120,0.0151591118839052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,51200,0.16263111432393393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,3584,0.013628444737858243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,5120,0.020353777541054618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,3072,0.012072888513406118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,4096,0.013403555585278405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,3584,0.01236355553070704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,4096,0.01759199963675605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,2560,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,3072,0.011049778097205691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,3584,0.016177778442700703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,16384,0.03305244445800781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,2048,0.008950221869680617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,1536,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,3072,0.01495911180973053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,2560,0.0101742222905159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,1024,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,2560,0.013385777672131857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,2048,0.008893333375453949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,768,0.00519466648499171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,1536,0.00853066643079122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,2048,0.01129422254032559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,10240,0.035535110367669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,512,0.0041946665280395085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,1536,0.009662222531106737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,256,0.0035031110876136353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,768,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,1024,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,512,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,128,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,768,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,64,0.002537777854336633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,256,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,32,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,128,0.00665155549844106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,128,0.006292444550328785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,64,0.006333333336644703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,65536,0.18869600031110975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,32,0.0064737776087390045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,51200,0.14801155196295843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,65536,0.1128657791349623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,16384,0.05046666661898295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,51200,0.08984800179799397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,51200,0.12666311528947619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,12288,0.039670222335391574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,16384,0.03236444460021125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,16384,0.04355466696951124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,1024,0.007698666718271043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,10240,0.03357955482270982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,12288,0.025470222036043804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,8192,0.02755911151568095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,12288,0.033906665113237165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,7168,0.02418577836619483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,10240,0.02267644471592373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,512,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,10240,0.028855111863878038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,6144,0.020763556162516277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,7168,0.018523555662896898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,5120,0.017824000782436795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,7168,0.022288888692855835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,6144,0.01623288955953386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,4096,0.014767110347747803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,6144,0.01991911066903008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,5120,0.014364444547229342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,3584,0.01312800000111262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,5120,0.017437333861986797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,4096,0.012437333663304647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,3072,0.011725333001878528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,65536,0.1594568888346354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,4096,0.014984889162911309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,2560,0.010290666586822933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,3584,0.012238221863905588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,3584,0.013904000322024027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,3072,0.01069155583779017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,2048,0.008947555389669206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,2560,0.009391110804345872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,3072,0.012859555582205454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,1536,0.0074942219588491656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,2560,0.011128000087208219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,2048,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,1024,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,2048,0.009712888962692684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,1536,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,768,0.004567999806669024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,8192,0.019426667028003268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,1536,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,1024,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,512,0.003866666720973121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,8192,0.024438222249348957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,256,0.0032097777972618737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,768,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,512,0.006991110742092133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,128,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,512,0.006404444575309753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,256,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,64,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,128,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,128,0.005983110931184556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,64,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,32,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,65536,0.13982578118642172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,65536,0.07106488943099976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,256,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,65536,0.15358755323621961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,1024,0.00737955586777793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,51200,0.11373955673641628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,16384,0.04299999939070808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,768,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,51200,0.05494755506515503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,12288,0.029663999875386555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,16384,0.02259911100069682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,256,0.006215999937719769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,10240,0.025825776987605627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,51200,0.12085510624779595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,16384,0.04195644458134969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,8192,0.021351110604074266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,12288,0.017596445149845548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,32,0.0028577778074476454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,7168,0.018183999591403537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,10240,0.015835555063353646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,12288,0.032315555546018816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,6144,0.015848888291252982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,8192,0.023843554986847773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,10240,0.02812444501452976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,7168,0.01275111072593265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,5120,0.013815999858909182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,4096,0.011811555259757571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,7168,0.02149066660139296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,6144,0.01934400035275353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,3584,0.010620444185203977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,4096,0.009323555562231276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,5120,0.01682488951418135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,4096,0.014576888746685453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,3072,0.00944622192117903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,2560,0.008656000097592672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,3584,0.013097777962684631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,3072,0.008291555775536431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,2048,0.00740444411834081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,3072,0.011876444849703046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,2560,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,1536,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,2048,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,2048,0.009360888765917884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,1024,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,1536,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,1536,0.008353778057628209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,8192,0.014097778333557976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,1024,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,6144,0.011967110964987012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,1024,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,512,0.0035751110149754416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,5120,0.010346666806273991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,768,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,768,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,256,0.0032142222755485107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,512,0.006235555642180973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,256,0.006028444402747684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,512,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,3584,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,64,0.0025235555238193935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,256,0.005976000179847081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,32,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,128,0.0061048890153567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,128,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,65536,0.12574666076236302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,2560,0.010475555227862464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,64,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,32,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,65536,0.06720355484220716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,51200,0.10133955876032512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,16384,0.04080355498525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,65536,0.1166986624399821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,12288,0.028647998968760174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,51200,0.053640001349978976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,10240,0.02385155525472429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,51200,0.0927928884824117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,16384,0.02092444399992625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,12288,0.017511111166742113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,16384,0.03193511234389411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,8192,0.0196435550848643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,12288,0.025703110628657874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,7168,0.01722577710946401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,8192,0.013790222505728403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,10240,0.02256444427702162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,8192,0.019259555472267997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,128,0.0028080000645584534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,7168,0.012608889076444836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,7168,0.017629333668284945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,5120,0.013441777891582914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,6144,0.011491555306646558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,4096,0.01145511120557785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,6144,0.015975111060672335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,5120,0.010087110930018956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,3584,0.010399111443095738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,4096,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,5120,0.014020444618331062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,4096,0.012317332956526013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,3072,0.009311111436949836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,3584,0.008623110751310984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,3584,0.011247110863526663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,2560,0.008285333712895712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,3072,0.008343111309740279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,2048,0.007732444339328342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,768,0.004233777936961916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,3072,0.010050666828950247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,2560,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,1536,0.006559111177921295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,2048,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,2560,0.00905866672595342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,10240,0.015638222297032673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,1536,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,2048,0.008294222255547842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,768,0.004221333397759331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,1536,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,6144,0.01532977819442749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,1024,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,512,0.0038000001675552795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,768,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,256,0.0029253332565228143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,1024,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,768,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,512,0.006016888966162999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,256,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,512,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,32,0.002472888885272874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,128,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,128,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,65536,0.10120266675949097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,32,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,51200,0.0779128869374593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,65536,0.06621955500708686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,65536,0.11683289210001628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,16384,0.031594667169782854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,1024,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,51200,0.052892443206575185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,12288,0.025801777839660645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,51200,0.08865244521035089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,16384,0.0206675562593672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,16384,0.03169599837727017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,10240,0.021798221601380244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,12288,0.017143110434214275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,128,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,8192,0.017542221479945712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,64,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,10240,0.015488000379668342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,7168,0.015923556354310777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,10240,0.022043555974960327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,8192,0.013064000341627332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,8192,0.018970666660202872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,7168,0.012069332930776807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,5120,0.012071111136012606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,6144,0.010608889162540436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,7168,0.017323555217848886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,4096,0.00998933364947637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,6144,0.015810666812790763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,5120,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,3584,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,5120,0.013823111024167804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,4096,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,3072,0.008023111356629265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,4096,0.011734222372372946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,3584,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,2560,0.007558222446176741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,3584,0.010737777584128909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,2048,0.006127999888526068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,2560,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,2560,0.008991999758614434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,1536,0.005114666703674528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,2048,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,2048,0.008244444098737504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,1024,0.004171555654870139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,12288,0.025567111041810777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,1536,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,768,0.0035386664999855887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,1536,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,1024,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,1024,0.006647999915811751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,512,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,768,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,768,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,256,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,512,0.006075555665625467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,512,0.005956444475385878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,128,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,256,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,256,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,3072,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,128,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,32,0.0025066667132907445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,128,0.005664888769388199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,64,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,32,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,6144,0.013908444179428948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,65536,0.11069689194361369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,64,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,65536,0.06442399819691975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,51200,0.07237955596711901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,16384,0.02999911043379042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,51200,0.051685333251953125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,16384,0.02015822297996945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,12288,0.024918221765094336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,51200,0.08652177784177993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,16384,0.031103110975689355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,10240,0.02016444504261017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,64,0.0024968888610601425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,12288,0.016897777716318767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,8192,0.016837333639462788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,3072,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,12288,0.02481866710715824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,7168,0.015133332875039844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,6144,0.013358222113715278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,8192,0.01294133315483729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,7168,0.011422221859296164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,65536,0.09286666578716701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,8192,0.018567111757066514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,5120,0.011704000333944956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,6144,0.010718222293588849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,6144,0.015433778365453085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,5120,0.009649777577983009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,3584,0.009069333473841349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,5120,0.013108444710572561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,4096,0.008771555291281806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,3072,0.008377778033415476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,4096,0.011125333607196808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,3584,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,3584,0.01035733355416192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,2560,0.007547555698288812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,3072,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,2048,0.006468444648716185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,3072,0.009735999835862054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,2560,0.007714666426181793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,1536,0.0053137776752312975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,2560,0.008974221845467886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,2048,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,1024,0.004229333251714706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,2048,0.008251555263996124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,1536,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,10240,0.015141333142916361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,768,0.0038328886859946777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,10240,0.021870222356584337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,1024,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,768,0.006313777632183499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,4096,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,768,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,256,0.0026871110830042097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,512,0.00608711110221015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,512,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,128,0.002479111154874166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,256,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,64,0.0022124445272816554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,128,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,256,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,32,0.0023164444913466773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,128,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,65536,0.07548177904552884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,64,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,32,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,51200,0.05775199996100532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,65536,0.10646488931443955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,16384,0.023426666855812073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,51200,0.050310220983293324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,1536,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,51200,0.08358577887217204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,12288,0.01863555610179901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,1024,0.006649777707126405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,16384,0.019911999503771465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,7168,0.01702311138312022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,10240,0.020785777105225455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,12288,0.016533333394262526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,16384,0.03007377849684821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,8192,0.017235555582576327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,7168,0.015503111812803479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,12288,0.024101333485709295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,10240,0.013639999760521783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,6144,0.01384800010257297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,10240,0.021105777886178758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,7168,0.011065777805116443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,7168,0.01646133263905843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,5120,0.011942221886581846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,6144,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,6144,0.014514666464593677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,4096,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,5120,0.009604444106419882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,65536,0.06351021925608318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,5120,0.012471111284361945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,3584,0.007702222300900354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,4096,0.008683555656009251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,3072,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,3584,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,512,0.0032088889016045462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,3584,0.010300444232092964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,2560,0.006380444599522485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,3072,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,2048,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,3072,0.009389333426952362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,2560,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,1536,0.004844444493452708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,2560,0.008791999684439765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,2048,0.007248889240953658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,8192,0.01239466667175293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,2048,0.00814844419558843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,1024,0.0035582222044467926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,768,0.0032195556494924757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,8192,0.018028444714016385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,1536,0.007242666350470648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,1024,0.0063635553750726916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,512,0.0028479999552170434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,768,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,1024,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,256,0.002543111021320025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,512,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,128,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,512,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,256,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,64,0.0022124445272816554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,256,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,128,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,32,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,4096,0.010713777608341642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,128,0.005283555636803309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,65536,0.05807822280459934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,64,0.0053031109273433685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,51200,0.04735555582576328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,32,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,65536,0.062124444378746875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,16384,0.01852888862291972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,65536,0.06667644447750516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,12288,0.01516533394654592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,51200,0.04892888996336195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,16384,0.019714666737450492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,51200,0.052744001150131226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,10240,0.015963556038008798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,16384,0.02034133341577318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,1536,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,8192,0.013617777989970313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,12288,0.01575199928548601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,7168,0.012258666257063547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,12288,0.016696888539526198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,10240,0.013451555536852943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,6144,0.010984000232484607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,8192,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,10240,0.01479644411140018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,7168,0.011077333655622272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,5120,0.010873777998818291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,768,0.006175999840100606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,8192,0.0127688886390792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,4096,0.009445333646403419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,6144,0.010395555860466426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,3584,0.008082666330867344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,7168,0.01204533295498954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,5120,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,3072,0.006393777827421824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,6144,0.0107342220014996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,4096,0.008674666285514832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,2560,0.00600000015563435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,5120,0.009723555710580613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,2048,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,3072,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,4096,0.008565333154466417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,1536,0.004206222378545337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,3584,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,2560,0.007702222300900354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,3072,0.007671111159854465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,1024,0.003563555578390757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,2048,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,2560,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,768,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,1536,0.006712889091836081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,512,0.0028320000403457214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,2048,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,1024,0.006401777681377199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,1536,0.006261333409282897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,768,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,1024,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,512,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,768,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,256,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,512,0.005454222361246745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,64,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,256,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,32,0.00222044438123703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,128,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,128,0.005142222262091107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,65536,0.04221066501405504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,64,0.005037333402368757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,51200,0.03341689043574863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,32,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,65536,0.06394222047593859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,16384,0.014426666829321118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,51200,0.047705776161617704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,51200,0.05011555552482605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,12288,0.012311110893885294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,16384,0.017679111825095285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,16384,0.01993688941001892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,256,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,10240,0.011764444410800934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,12288,0.01479288935661316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,128,0.0021946666141351065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,12288,0.01586844523747762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,8192,0.010636444720957013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,10240,0.013463111387358772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,7168,0.011339555184046427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,8192,0.011759999725553723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,6144,0.010017777482668558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,7168,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,8192,0.01236355553070704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,5120,0.008807111117574904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,7168,0.011409777734014722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,6144,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,4096,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,6144,0.010655110908879174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,65536,0.05975821945402357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,5120,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,3584,0.006554666492674086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,5120,0.009320889082219865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,3072,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,3584,0.008447111480765873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,4096,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,2560,0.005203555441564984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,3584,0.008310221963458592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,2048,0.004549333204825719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,3072,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,3584,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,2560,0.007626666790909237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,1536,0.003903111235962974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,3072,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,1024,0.003175111073586676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,2048,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,1536,0.006741333338949416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,2048,0.006704000135262807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,768,0.0028320000403457214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,1024,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,1536,0.006212444355090459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,512,0.0026764445420768526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,768,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,1024,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,256,0.00245066670080026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,768,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,10240,0.013985777894655863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,128,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,512,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,256,0.005744888964626525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,64,0.0019022222194406723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,32,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,256,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,65536,0.03230488962597317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,128,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,64,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,51200,0.02678933408525255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,65536,0.05795466899871826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,4096,0.008292444050312042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,65536,0.06256977717081706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,16384,0.012845333251688214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,51200,0.04667733444107903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,51200,0.049136890305413135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,2560,0.007132444116804335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,16384,0.017495110630989075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,10240,0.010129777921570672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,16384,0.019170666734377544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,12288,0.014759999182489185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,8192,0.009073778159088558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,12288,0.015512888630231222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,10240,0.013186666700575085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,7168,0.00998577806684706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,512,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,10240,0.013803555733627744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,8192,0.011708444191349877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,8192,0.012069332930776807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,7168,0.01093066649304496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,128,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,5120,0.008307555483447181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,7168,0.011352889239788055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,6144,0.010092444717884064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,4096,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,5120,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,32,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,5120,0.009284444153308868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,4096,0.008655110994974772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,4096,0.00833511104186376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,3072,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,3584,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,12288,0.011243555280897351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,3584,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,2560,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,3072,0.00794666674402025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,2048,0.004550222307443619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,3072,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,2560,0.007351111206743453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,1536,0.003882666842805015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,2560,0.007056000332037608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,2048,0.007079111205206976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,1024,0.0032008888406885993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,2048,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,6144,0.009746666583749983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,1536,0.006683555742104848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,768,0.00281866660548581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,1536,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,512,0.002552888873550627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,768,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,1024,0.0057653333577844836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,6144,0.010332444475756751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,256,0.002313777804374695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,768,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,512,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,3584,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,512,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,256,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,64,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,32,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,256,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,128,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,65536,0.02490666674243079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,128,0.004977777600288391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,51200,0.021943110558721755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,64,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,32,0.004965333475006951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,65536,0.05887999799516466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,16384,0.013134222063753339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,65536,0.061108443472120494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,51200,0.0462355547481113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,12288,0.012527999778588613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,51200,0.04827910992834303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,16384,0.0173999998304579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,10240,0.011188444164064197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,16384,0.018601778480741713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,12288,0.01458666721979777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,8192,0.009541333549552495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,12288,0.015225778023401896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,10240,0.013418667018413544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,7168,0.008788444101810455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,1024,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,8192,0.011427555647161273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,10240,0.01350488927629259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,6144,0.008071111308203803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,8192,0.011905777785513135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,5120,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,7168,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,4096,0.006894222150246303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,6144,0.010076444182131026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,7168,0.011042666931947073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,3584,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,5120,0.009322666459613377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,6144,0.01034222212102678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,3072,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,4096,0.008398222426573435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,5120,0.009310222334331935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,3584,0.008290666672918532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,4096,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,3072,0.007675555845101674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,2048,0.004517333375083076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,3584,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,1536,0.0038613333470291565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,3072,0.007626666790909237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,2560,0.0074924445814556545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,1024,0.0031715554909573663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,2048,0.0070053330726093715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,1536,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,2048,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,768,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,1024,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,1536,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,512,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,1024,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,768,0.006230222268237009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,256,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,512,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,768,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,128,0.0021155555215146807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,512,0.005321777943107817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,256,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,64,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,32,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,128,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,65536,0.02074666652414534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,768,128,0.002190222135848469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,128,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,64,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,32,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,51200,0.01700622174474928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,65536,0.05396533343527052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,512,2560,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,16384,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,51200,0.04343466626273262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,12288,0.009369778136412302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,16384,0.01720622181892395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,51200,0.046329776446024575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,2560,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,10240,0.008767111433876885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,16384,0.018366222580273945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,8192,0.009082666701740688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,12288,0.014344889256689282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,10240,0.013139555851618448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,12288,0.01482222146458096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,7168,0.008344000412358178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,8192,0.011377777490350934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,6144,0.00793422261873881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,8192,0.011751111182901593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,7168,0.01073333372672399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,5120,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,7168,0.011032889286677042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,256,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,6144,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,4096,0.006700444552633498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,6144,0.010088000032636855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,3584,0.006230222268237009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,4096,0.008291555775536431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,5120,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,3072,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,4096,0.008037333687146505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,3584,0.008020444048775567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,2560,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,3584,0.007888000044557784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,3072,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,2048,0.004348444441954295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,3072,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,2560,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,1536,0.003738666574160258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,2560,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,2048,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,1024,0.003181333343187968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,10240,0.013408888545301227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,1536,0.005972444597217772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,768,0.0028640000770489373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,1024,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,1024,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,512,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,768,0.006230222268237009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,768,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,5120,0.009332444104883406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,256,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,512,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,512,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,128,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,65536,0.057872891426086426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,128,0.0058666666348775225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,256,32,0.0017884444031450485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,256,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,65536,0.015267555912335714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,128,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,51200,0.013055110971132914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,64,0.004968889057636261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,32,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,65536,0.05126755436261495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,16384,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,65536,0.05750222338570488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,2048,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,12288,0.011381333072980246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,1536,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,51200,0.04145955708291795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,10240,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,16384,0.01720888912677765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,51200,0.045778665277693004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,8192,0.008938666847017076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,12288,0.014451555079884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,10240,0.013092444174819522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,12288,0.014862222803963555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,8192,0.011400889191362592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,10240,0.013343999783198038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,6144,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,8192,0.011415111521879831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,7168,0.010741333166758219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,5120,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,6144,0.00999644481473499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,7168,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,4096,0.006555555595291986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,6144,0.010020444790522257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,256,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,5120,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,3584,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,3072,0.005404444618357553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,5120,0.008764444126023186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,3584,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,4096,0.007782222496138678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,2560,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,3072,0.007856888903511895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,3584,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,2048,0.0042364444169733255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,3072,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,2560,0.007342221836249034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,2560,0.006858666737874349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,1536,0.003621333175235324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,2048,0.007139555282062954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,1024,0.0032053333189752367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,1536,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,16384,0.018214222457673814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,768,0.0028115556471877625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,7168,0.008349333372380998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,1024,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,512,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,1024,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,768,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,256,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,768,0.005305777821275923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,512,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,128,0.001998222122589747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,512,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,256,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,64,0.0018311111877361934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,256,0.005102222164471944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,128,32,0.0018008889423476325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,128,0.005666666560702854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,128,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,65536,0.01909955508179135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,51200,0.01686844395266639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,64,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,16384,0.00830844458606508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,32,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,12288,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,65536,0.05682577689488729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,51200,0.04526844289567736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,2048,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,16384,0.018035555879275005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,1536,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,12288,0.01478488908873664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,7168,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,6144,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,10240,0.01311644415060679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,5120,0.006330666856633292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,4096,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,8192,0.011626666618718041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,3584,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,7168,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,6144,0.009769777456919352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,3072,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,2560,0.004555555681387584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,5120,0.009038222332795462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,4096,0.008392888638708327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,2048,0.004265777766704559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,1536,0.003579555700222651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,4096,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,1024,0.0028488888508743713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,3072,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,768,0.0028204443968004654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,2560,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,2048,0.006406222366624408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,1536,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,256,0.002201777779393726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,1024,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,128,0.0018800000349680583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,10240,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,768,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,512,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,32,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,8192,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,65536,0.020634666085243225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,256,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,51200,0.01835466590192583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,128,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,16384,0.014797333214018079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,64,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,12288,0.011808888779746162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,32,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,10240,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,8192,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,65536,0.05647022194332547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,51200,0.04521155688497755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,7168,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,6144,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,12288,0.014829332629839579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,5120,0.006310222049554189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,10240,0.013166222307417126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,4096,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,8192,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,3584,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,3584,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,6144,0.010007999837398529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,5120,0.008815999660227034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,64,512,0.002492444382773505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,2560,0.00461422238085005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,4096,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,2048,0.004192000048028098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,3584,0.007789333661397298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,3072,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,2560,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,2048,0.006561777657932705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,1024,0.002874666617976295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,768,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,1536,0.006161777923504512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,512,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,1024,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,256,0.0021057778762446507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,768,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,128,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,512,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,64,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,16384,0.018191999859280057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,32,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,256,0.005060444275538127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,128,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,7168,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,64,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,3072,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,32,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,257,32,1536,0.003504000190231535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,16384,0.4931519826253255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,16384,0.34528088569641113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,12288,0.3781386746300592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,16384,0.14470667309231228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,51200,1.07787291208903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,51200,1.5264239841037328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,51200,0.4291697873009576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,12288,0.25852976904975045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,12288,0.1107448869281345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,8192,0.20953422122531465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,10240,0.09139111306932236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,16384,0.31985865698920357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,8192,0.17796622382269967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,8192,0.07573778099483915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,7168,0.17986933390299478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,12288,0.24312710762023926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,10240,0.20075289408365884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,7168,0.15926222006479898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,7168,0.06691022051705255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,6144,0.15254933304256862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,8192,0.16247733434041342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,10240,0.2692355579800076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,5120,0.1304595602883233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,6144,0.1331662204530504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,6144,0.057928886678483754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,7168,0.14738133218553331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,4096,0.10934222406811184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,5120,0.11399555206298828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,5120,0.04976266622543335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,3584,0.09369688563876682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,4096,0.042007999287711255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,4096,0.09502222140630086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,6144,0.12626132700178358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,5120,0.1049013336499532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,3072,0.09102577633327907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,3584,0.037680887513690524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,4096,0.0865137775739034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,3584,0.08023199770185682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,3072,0.06920799944135878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,3584,0.07712888717651367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,2560,0.060489780373043485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,3072,0.06820711162355211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,2048,0.06144977940453423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,2560,0.029634667767418757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,2560,0.058563556936052114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,2048,0.02574399941497379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,2048,0.05116711060206095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,1536,0.0435733331574334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,2048,0.05008800162209404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,1536,0.040463109811147056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,10240,0.2486168808407254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,1536,0.020596444606781006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,1024,0.03303555647532145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,768,0.026258667310078938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,1536,0.04205422268973457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,1024,0.015809777710172866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,2560,0.06807377603318956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,1024,0.03200533323817783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,512,0.018602665927675035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,3072,0.03362133436732822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,768,0.025999110605981614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,768,0.014254222313563028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,256,0.011211555865075854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,51200,0.9500968721177844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,768,0.030093332131703694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,512,0.02156977852185567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,128,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,256,0.01906577746073405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,1024,0.03507466779814826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,256,0.01516888870133294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,128,0.01831822262869941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,128,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,512,0.013075555364290873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,32,0.008489777644475302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,512,0.026367111338509455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,128,0.01720622181892395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,64,0.017972444494565327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,32,0.017885333961910672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,256,0.021196444829305012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,64,0.00849866701496972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,65536,1.076708475748698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,65536,1.356056849161784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,51200,0.8576035499572754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,65536,0.4362782107459174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,16384,0.2737102243635389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,16384,0.3389608860015869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,12288,0.25304799609714085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,16384,0.11332355605231391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,51200,0.3369990984598796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,51200,1.0463484658135307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,12288,0.20693066385057238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,10240,0.18027822176615396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,12288,0.08572177754508124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,10240,0.16980977853139242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,8192,0.1471066607369317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,10240,0.07262044482760958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,16384,0.2604915565914578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,8192,0.06002133422427707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,8192,0.13761511113908556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,12288,0.1947235531277127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,7168,0.1558862262301975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,10240,0.16346577803293863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,6144,0.11082133319642808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,7168,0.11965244346194798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,7168,0.05274933245446947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,8192,0.13336888949076334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,5120,0.09544533491134644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,6144,0.10651733477910359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,6144,0.04710844490263197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,4096,0.07834577560424805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,5120,0.03963022099600898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,6144,0.10224355591668023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,5120,0.08617689212163289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,4096,0.07240177525414361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,3584,0.06708977619806926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,4096,0.07185333304935031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,3584,0.06493155823813544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,3584,0.06347733073764376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,3072,0.058387557665506996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,5120,0.09013599819607204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,3584,0.030470222234725952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,2560,0.055797331862979464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,3072,0.027383999692069158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,3072,0.055460443099339805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,51200,0.7941368950737847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,2048,0.0399875541528066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,4096,0.034133334954579674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,65536,1.0039360258314345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,2560,0.024463999602529738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,3072,0.05800888935724894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,1536,0.031742221779293485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,2048,0.023912888434198167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,2560,0.05029333300060696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,2048,0.04156800111134847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,1024,0.024113777610990737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,1536,0.033322665426466204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,1536,0.019973332683245342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,2048,0.043290668063693576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,768,0.019996444384257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,1024,0.013847110999955071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,1536,0.03714222378200955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,1024,0.026778666509522334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,512,0.01330044451687071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,7168,0.11808889442019993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,768,0.012854222622182635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,768,0.021273778544531927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,256,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,1024,0.030893332428402368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,512,0.011707555916574268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,768,0.027648889356189307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,128,0.008787555827034844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,256,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,512,0.02275733318593767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,256,0.01611199975013733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,128,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,2560,0.04885511265860664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,128,0.015254222684436373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,32,0.008681777450773451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,256,0.01923111081123352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,128,0.015229332778188916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,64,0.01514755520555708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,32,0.014871999621391296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,65536,0.39188443289862734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,51200,0.3065155612097846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,65536,0.14790933661990696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,51200,0.2895439995659722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,16384,0.11368710464901394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,512,0.017550221747822232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,51200,0.11730399396684434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,64,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,16384,0.04116888840993246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,16384,0.09364977810117935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,12288,0.08684000041749741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,10240,0.06495644648869832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,12288,0.03267733256022135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,16384,0.09098044368955825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,12288,0.07079111205206977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,8192,0.05424977673424614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,10240,0.028184887435701158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,10240,0.05996888875961304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,65536,0.36005067825317383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,12288,0.07179466883341472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,8192,0.049716442823410034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,8192,0.023677332533730402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,7168,0.047671112749311656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,10240,0.062099556128184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,65536,0.3413742118411594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,6144,0.04455022348297966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,8192,0.05165955424308777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,7168,0.021538666552967493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,7168,0.04373066624005636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,5120,0.03596533338228861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,6144,0.019175999694400363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,6144,0.038663112454944186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,7168,0.04740088846948412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,4096,0.030594666798909504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,6144,0.042797333664364286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,5120,0.016833777228991192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,5120,0.03304444419013129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,3584,0.027621333797772724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,4096,0.027991112735536363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,4096,0.015006222658687167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,3072,0.023941333095232647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,3584,0.024775111012988623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,4096,0.032808889945348106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,3584,0.013832888669437833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,2560,0.020036444067955017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,3072,0.013046222428480784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,3584,0.0299253331290351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,3072,0.022454222043355305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,51200,0.26740532451205784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,2560,0.012030222349696688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,3072,0.02798844377199809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,1536,0.0127697777416971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,2048,0.011036444041464063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,2560,0.02507466740078396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,2048,0.01717155509524875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,2048,0.02197244432237413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,1024,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,1536,0.014316444595654806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,1536,0.008969777988062965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,768,0.008623110751310984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,5120,0.037789334853490196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,1024,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,1024,0.011777777638700275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,512,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,768,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,1024,0.016916443904240925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,2048,0.0162453336848153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,512,0.009014222357008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,768,0.01551999979548984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,256,0.004584889031118817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,2560,0.019540444016456604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,128,0.0042088888585567474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,256,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,512,0.01310488912794325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,256,0.008396444221337637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,64,0.003945777813593547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,128,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,256,0.012060444388124677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,32,0.0041991112132867174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,128,0.008353778057628209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,128,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,32,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,65536,0.2942417727576362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,1536,0.01941511034965515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,768,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,65536,0.29305956098768443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,51200,0.2308702203962538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,512,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,65536,0.1298204395506117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,51200,0.230059544245402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,16384,0.08026844263076782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,51200,0.10203466812769572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,16384,0.03855822152561612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,16384,0.07781510882907443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,12288,0.05927199787563748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,12288,0.030481778913074072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,16384,0.08847555849287246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,10240,0.05127822359402975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,12288,0.0685831109682719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,10240,0.025622222158643935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,10240,0.05106488863627116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,8192,0.04485599862204658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,65536,0.3196053240034315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,51200,0.25161245134141713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,8192,0.021926222576035395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,10240,0.06046577956941393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,8192,0.04244977898067898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,6144,0.03220444586541917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,7168,0.01997866729895274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,8192,0.05133777856826782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,7168,0.03736622134844462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,12288,0.06194755766126844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,5120,0.02776977750990126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,6144,0.03290133343802558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,7168,0.04547199938032362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,4096,0.02422222163942125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,5120,0.028583112690183852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,5120,0.01621955633163452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,6144,0.041442665788862444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,64,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,3584,0.02107644412252638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,5120,0.03650488787227207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,4096,0.023807111713621352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,4096,0.014465777410401238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,3072,0.01815111107296414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,3584,0.013393777940008374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,3584,0.02142133315404256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,4096,0.03180177675353156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,7168,0.03933155536651611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,2560,0.01513066722287072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,3072,0.0125928885406918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,3584,0.029106666644414265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,2048,0.013063111239009433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,2560,0.011642666326628791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,3072,0.026405332816971674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,2560,0.017299556069903903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,1536,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,6144,0.01794311073091295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,2560,0.024123556084103052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,2048,0.015000000596046448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,1024,0.008735111190213097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,1536,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,1536,0.012773333324326409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,2048,0.021505778034528095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,768,0.007589333587222629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,1024,0.007696000238259633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,1024,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,1536,0.018995554910765756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,512,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,768,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,1024,0.016050666570663452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,3072,0.019347555107540555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,512,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,256,0.004567111117972268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,512,0.0069919998447100324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,768,0.014150222142537435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,128,0.0038808890514903595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,256,0.006670222100284364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,512,0.01294133315483729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,256,0.007344000041484833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,64,0.003826666623353958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,128,0.007337777978844113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,2048,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,128,0.006686222222116258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,32,0.003995555556482739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,256,0.012050666742854647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,64,0.007336888876226213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,128,0.009666666388511658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,65536,0.2305662102169461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,32,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,51200,0.18358666367001006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,65536,0.25119821230570477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,768,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,16384,0.06160177787144979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,65536,0.12538666195339626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,51200,0.19506933954026964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,12288,0.05249244305822584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,16384,0.037061333656311035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,16384,0.06678489181730482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,10240,0.041165331999460854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,12288,0.05093422200944689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,12288,0.028591109646691218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,16384,0.08699466784795125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,8192,0.034061332543691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,10240,0.024604444702466328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,12288,0.0683786670366923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,51200,0.24870488378736708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,65536,0.3185848924848768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,8192,0.02124533388349745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,10240,0.05926577912436592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,8192,0.03673511081271701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,6144,0.02574933403068119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,7168,0.019339554839664035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,8192,0.050083554453319974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,7168,0.032574223147498235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,51200,0.09981599781248306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,5120,0.02288444505797492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,6144,0.02849510974354214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,6144,0.017535999417304993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,7168,0.045556445916493736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,4096,0.01944977707333035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,5120,0.024708444873491924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,5120,0.015807999504937064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,3584,0.01609422266483307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,6144,0.0409457782904307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,4096,0.020851555797788832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,4096,0.014195554786258273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,5120,0.03601955705218845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,3072,0.014864888456132678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,10240,0.04393066631423103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,3584,0.013055110971132914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,4096,0.030600888861550227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,3584,0.018772444791264005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,3072,0.011976000335481433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,3072,0.017329777280489605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,2048,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,7168,0.03012088934580485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,3584,0.02811288833618164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,2560,0.0107004443804423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,2560,0.015491555134455362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,3072,0.0255084451701906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,1536,0.008867555194430882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,2048,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,2560,0.023358222511079576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,1024,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,1536,0.011694221860832639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,1536,0.008160889148712158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,2048,0.021091555555661518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,768,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,1536,0.0185217774576611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,1024,0.009321778184837764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,768,0.007276444799370236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,768,0.008359111017651027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,512,0.0047315553658538395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,1024,0.01499644418557485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,2560,0.01237422227859497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,512,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,768,0.01387555566098955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,256,0.003625777860482534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,512,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,128,0.0032017777363459268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,256,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,512,0.012735111018021902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,256,0.0070062221752272705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,64,0.0031679999083280563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,2048,0.013625777430004544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,128,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,256,0.011630222201347351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,64,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,128,0.00960355583164427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,65536,0.21140978071424696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,32,0.007249777515729268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,1024,0.007667555577225155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,51200,0.15703999996185303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,65536,0.2071004443698459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,65536,0.07865689198176067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,51200,0.1605795489417182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,16384,0.058371557129753955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,51200,0.06321689155366686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,16384,0.05550133188565572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,128,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,16384,0.02405066622628106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,12288,0.04260266820589701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,16384,0.053991112444135875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,10240,0.03546222382121616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,65536,0.18188088470035127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,51200,0.14647732840643987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,12288,0.019444444113307528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,8192,0.030518223841985066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,12288,0.04285688863860237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,10240,0.016992888516849942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,10240,0.0370399985048506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,7168,0.026851556367344324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,12288,0.04289244280921089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,8192,0.015048889650238885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,8192,0.030880000856187608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,6144,0.022963555322753057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,10240,0.03827822208404541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,7168,0.027186666925748188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,7168,0.013822221921549903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,8192,0.03269422385427687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,6144,0.024318221542570326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,7168,0.03006311257680257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,4096,0.016484444340070088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,5120,0.02089511189195845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,5120,0.011693333586057028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,6144,0.02680533296532101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,3584,0.014314666390419006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,4096,0.010682666467295753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,5120,0.02381244467364417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,4096,0.018061333232455783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,3072,0.012833777401182385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,3584,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,3584,0.01650399963061015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,4096,0.021680000755521987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,2560,0.011062222222487131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,3072,0.015000889698664347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,3072,0.008926221893893348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,3584,0.020059555768966675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,2048,0.009678222239017487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,2560,0.013766222529941134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,2560,0.008040889269775814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,3072,0.01886311173439026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,1536,0.0074551113777690465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,5120,0.019859555694792006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,2048,0.01204622205760744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,2048,0.007584888901975419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,6144,0.012825777961148156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,1024,0.006227555374304454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,2560,0.017743999759356182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,1536,0.007009777757856581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,1536,0.010330666270520952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,768,0.005126222140259213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,2048,0.016526222229003906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,1024,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,512,0.00406577769252989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,1024,0.008371555142932469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,1536,0.013893332746293811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,32,0.003481777591837777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,768,0.006501333167155583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,256,0.0032115555885765287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,768,0.007620444728268518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,1024,0.012098666694429187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,512,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,512,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,128,0.002975111206372579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,768,0.011401777466138205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,256,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,256,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,64,0.0028533333291610083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,32,0.0031911111954185697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,128,0.006005333529578314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,128,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,256,0.01035377797153261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,65536,0.17812889152102998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,64,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,128,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,32,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,51200,0.14497244358062744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,65536,0.20044178432888457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,65536,0.07465689049826728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,16384,0.04761688907941183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,51200,0.15631200207604304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,51200,0.059454222520192467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,12288,0.04074666566318936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,16384,0.02331466641690996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,16384,0.05376088950369093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,51200,0.1354400051964654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,65536,0.17426221900516084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,12288,0.04165510998831855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,16384,0.05270933442645603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,10240,0.03163466519779629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,512,0.010787555740939246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,8192,0.026752000053723652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,10240,0.016590222716331482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,10240,0.03571644425392151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,12288,0.04251911242802938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,7168,0.023423110445340473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,8192,0.029618667231665716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,8192,0.014467555615637036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,10240,0.037384000089433454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,7168,0.013570666313171387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,8192,0.031940443648232356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,7168,0.02679466704527537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,5120,0.017163554827372234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,6144,0.012450666891203986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,7168,0.02918222215440538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,5120,0.02035555574629042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,5120,0.011371555427710215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,6144,0.026526222626368206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,4096,0.014635556273990206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,3584,0.01311377767059538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,4096,0.010235555469989777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,5120,0.023659555448426142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,4096,0.01756888959142897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,3072,0.011902222202883827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,3584,0.009160888691743216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,4096,0.02119111186928219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,3584,0.016134222348531086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,2560,0.01038933296998342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,3072,0.008691555923885768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,3072,0.0147733340660731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,3584,0.02000177734427982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,2048,0.008839999636014303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,2560,0.013190222283204397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,2560,0.007876444194051955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,3072,0.018560889694425795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,1536,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,2048,0.007384888827800751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,2048,0.011630222201347351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,2560,0.017362667454613578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,1024,0.006112888869312074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,1536,0.009986666341622671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,1536,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,2048,0.016174222032229107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,768,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,12288,0.018610666195551556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,1024,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,1024,0.008169777691364288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,512,0.004211555752489301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,1536,0.013800000150998434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,768,0.006518222391605377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,768,0.007425777614116669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,256,0.003548444559176763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,1024,0.011787555283970304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,512,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,512,0.006904888898134232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,768,0.011067555182509951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,6144,0.020259555843141343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,128,0.0031946665710873077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,512,0.011149333582984077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,256,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,128,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,64,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,32,0.0032017777363459268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,128,0.006298666612969504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,256,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,6144,0.023500444160567388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,65536,0.17012977600097656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,128,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,32,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,51200,0.12900533941056994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,65536,0.16270489162868923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,65536,0.07141244411468506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,16384,0.04797688788837857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,51200,0.12612799803415933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,51200,0.05772978067398071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,65536,0.16487733523050943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,16384,0.04444800151718987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,16384,0.021943999661339655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,256,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,51200,0.1327164438035753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,16384,0.051573332813051015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,10240,0.02926666537920634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,12288,0.03456177645259433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,12288,0.018175111876593698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,8192,0.025454221500290766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,64,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,10240,0.01607111096382141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,10240,0.02942844563060337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,7168,0.02161866592036353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,8192,0.014150222142537435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,10240,0.03699822227160136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,7168,0.013173333472675748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,7168,0.022104000051816303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,8192,0.03184088733461168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,6144,0.018834667073355783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,6144,0.01204533295498954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,12288,0.03646311163902283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,5120,0.015995555453830294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,6144,0.019846222466892667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,6144,0.025977777110205755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,5120,0.01090844472249349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,4096,0.013737777868906656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,5120,0.017481777403089736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,3584,0.012076444096035428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,4096,0.009820444716347588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,5120,0.02346933384736379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,12288,0.04106400079197354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,4096,0.015303999185562134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,3072,0.010844444235165914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,3584,0.009023110899660323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,3584,0.01407199932469262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,4096,0.02084622283776601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,8192,0.024903110331959192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,2560,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,3072,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,3072,0.012896888785892062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,2048,0.008292444050312042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,2560,0.00794044468137953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,3584,0.019889778561062284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,1536,0.007055111229419708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,2560,0.011708444191349877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,2048,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,7168,0.02861066659291585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,2048,0.009909333454238044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,1024,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,1536,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,2048,0.015825778245925903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,768,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,1024,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,1024,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,1024,0.01205599970287747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,768,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,768,0.0069013333155049225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,512,0.0038977778620190094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,512,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,768,0.011565333439244164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,256,0.0032053333189752367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,512,0.006414222220579783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,512,0.01072533345884747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,256,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,3072,0.018588445252842374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,256,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,256,0.009736888938479954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,128,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,2560,0.017200888858901132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,64,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,128,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,1536,0.008348444269763099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,128,0.00793511089351442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,65536,0.14370400375790066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,64,0.005972444597217772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,1536,0.013479111095269522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,32,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,51200,0.12295111020406087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,65536,0.15546488761901855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,16384,0.0389395554860433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,65536,0.06840622425079346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,51200,0.12172533406151666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,51200,0.05508888761202494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,12288,0.032207111517588295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,16384,0.020933333370420668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,16384,0.04298488961325752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,128,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,65536,0.1638960043589274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,10240,0.025713778204388086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,16384,0.05106044477886624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,12288,0.017376888129446242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,32,0.0031155554784668815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,8192,0.022292445103327434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,10240,0.015535111228624979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,10240,0.02865333358446757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,12288,0.04105777872933282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,7168,0.0188719994492001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,8192,0.013748444616794586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,10240,0.036179555786980525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,8192,0.024267555938826665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,6144,0.01685333251953125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,7168,0.021653332644038733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,7168,0.012841777669058906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,8192,0.030384000804689195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,5120,0.014415999253590902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,6144,0.011694221860832639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,6144,0.019768888751665752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,7168,0.028410666518741187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,4096,0.012330667012267642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,51200,0.13158044550153944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,5120,0.010442666709423065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,5120,0.017095999585257638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,3584,0.011183111204041375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,6144,0.025536000728607178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,12288,0.0329786671532525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,4096,0.009445333646403419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,3072,0.010031999813185798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,5120,0.023025777604844835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,3584,0.008727110922336578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,3584,0.013631111217869652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,4096,0.020789333515697055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,2560,0.008593777815500895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,3072,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,3072,0.012425777812798819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,3584,0.019503111640612285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,2048,0.007690666450394525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,2560,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,3072,0.018351111147138808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,2560,0.01109955542617374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,1536,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,2048,0.007335111498832703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,2560,0.016591999265882704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,1024,0.005029333134492238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,1536,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,1536,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,2048,0.014913777510325113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,768,0.0039057777159743835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,1024,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,1024,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,4096,0.014937778313954672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,768,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,768,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,512,0.003488000068399641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,1024,0.011776000261306763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,256,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,512,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,768,0.011506666739781698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,512,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,128,0.0028355556229750314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,256,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,512,0.0107342220014996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,64,0.002528888897763358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,128,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,256,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,128,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,32,0.0028124445428450904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,2048,0.009552888572216034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,65536,0.11576888296339248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,128,0.008365333080291748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,64,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,32,0.006042666733264923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,51200,0.08715644147660996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,1536,0.013397333522637686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,65536,0.06610577636294894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,16384,0.036835554573271014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,51200,0.05281955666012234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,51200,0.09114755524529351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,12288,0.028636445601781208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,16384,0.032672888702816434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,16384,0.020280889338917203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,51200,0.12457422415415446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,10240,0.02474222249454922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,12288,0.017112889223628573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,12288,0.025507556067572698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,10240,0.015153777268197803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,10240,0.022504889302783545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,12288,0.039120889372295804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,8192,0.01997244523631202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,8192,0.013238222234778933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,10240,0.03442044390572442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,7168,0.017659554878870647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,8192,0.01918044520748986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,65536,0.11876000298394097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,6144,0.015360888507631091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,8192,0.029245333539115057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,7168,0.017483555608325534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,65536,0.15593243969811335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,6144,0.011062222222487131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,7168,0.0266311119000117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,6144,0.024488000406159297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,5120,0.014000000225173103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,5120,0.009739555418491364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,4096,0.01036711119943195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,5120,0.022488888767030504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,4096,0.012120000190205045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,7168,0.012411555482281579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,3584,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,4096,0.008767999708652496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,4096,0.01998044384850396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,5120,0.013416888813177744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,3072,0.008295111358165741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,3584,0.011477332976129321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,3584,0.008387555678685507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,256,0.005981333139869902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,2560,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,3584,0.018722666634453666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,3072,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,3072,0.010050666828950247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,2048,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,2560,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,2560,0.008990222381220924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,3072,0.01719111038578881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,1536,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,2048,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,2048,0.007256000406212277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,2560,0.016345777445369296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,1024,0.004179555508825514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,1536,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,1536,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,768,0.003768889026509391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,2048,0.014154665999942355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,1024,0.00665511108107037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,1536,0.012796444197495779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,512,0.0032702222880389956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,6144,0.016048888365427654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,768,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,1024,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,256,0.0028773333049482773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,768,0.010801778071456485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,16384,0.047839999198913574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,512,0.005977777971161737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,128,0.002847111059559716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,256,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,256,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,512,0.010676444404655032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,64,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,128,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,32,0.0028213332924577924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,256,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,65536,0.10229244497087266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,128,0.0074035558435651995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,32,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,1024,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,65536,0.06505866845448811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,65536,0.11744977368248834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,51200,0.07998044623268975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,768,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,16384,0.03294666608174642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,512,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,65536,0.15416444672478571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,51200,0.05202844407823351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,51200,0.0913955569267273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,12288,0.028012444575627644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,16384,0.03230844603644477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,128,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,51200,0.12365689542558457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,10240,0.023600000474188063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,12288,0.01666844387849172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,12288,0.025799999634424847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,16384,0.04775377776887682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,8192,0.019175110591782462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,64,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,10240,0.015146666102939181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,12288,0.03801600138346354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,10240,0.022278222772810195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,8192,0.019146665930747986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,8192,0.01315822203954061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,6144,0.015106666419241162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,10240,0.03385688861211141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,7168,0.017430222696728177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,7168,0.011759999725553723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,5120,0.013062222136391534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,7168,0.026551999979548987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,6144,0.0158915552828047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,6144,0.010671111444632212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,4096,0.010062221851613786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,5120,0.009341333475377824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,5120,0.013927110367351107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,6144,0.024195555183622573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,3584,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,4096,0.008727110922336578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,4096,0.011737777955002256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,5120,0.022109333011839125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,3072,0.008031110796663497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,3584,0.011258666714032492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,4096,0.01962577799956004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,7168,0.01702844434314304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,2560,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,3072,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,8192,0.028556442923016016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,3072,0.01661600006951226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,2560,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,2048,0.006308444258239534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,2560,0.0076657773719893555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,2048,0.008248888783984715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,16384,0.02018311123053233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,2560,0.015346666177113851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,1536,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,2048,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,2048,0.013954665925767688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,1024,0.0042168887125121225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,3584,0.008348444269763099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,1536,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,1536,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,768,0.003871110992299186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,1024,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,1536,0.0124177775449223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,3072,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,512,0.003548444559176763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,768,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,768,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,1024,0.011352000137170156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,256,0.002887111157178879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,512,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,512,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,768,0.010694222317801582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,128,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,256,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,256,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,512,0.010670222342014313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,128,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,128,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,32,0.0028640000770489373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,256,0.009257777697510189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,128,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,65536,0.09871911340289646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,64,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,51200,0.07363022036022611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,3584,0.018439110782411363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,65536,0.06394844584994845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,51200,0.0867768857214186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,51200,0.05127200153138903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,16384,0.030018667380015056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,65536,0.1543982161415948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,51200,0.12296977308061387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,16384,0.030993776188956365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,12288,0.021356445219781663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,64,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,16384,0.04775466521581014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,12288,0.016527111331621807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,10240,0.017324444320466783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,12288,0.024875554773542616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,32,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,12288,0.03816533419820998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,10240,0.014717333846622042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,8192,0.015201777219772339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,10240,0.021888888544506494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,1024,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,7168,0.013312000367376538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,8192,0.012376888758606382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,10240,0.03346666693687439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,8192,0.018602665927675035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,6144,0.011694221860832639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,7168,0.011050666371981302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,7168,0.017009778155220878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,5120,0.010527110762066312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,8192,0.028527110815048218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,6144,0.010072000324726105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,6144,0.015488889482286243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,7168,0.026212443908055622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,16384,0.01994222237004174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,5120,0.009383999639087254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,6144,0.024168888727823894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,3584,0.008093333078755273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,4096,0.008620444271299574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,4096,0.011343110766675739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,5120,0.021703110800849065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,3072,0.007520889242490132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,3584,0.01069866700304879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,4096,0.018816888332366943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,65536,0.11006755299038357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,3072,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,2560,0.006671999891599019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,3584,0.01789955629242791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,2560,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,3072,0.01688888834582435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,2048,0.005867555737495422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,2560,0.014582221706708273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,4096,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,2048,0.008013333711359236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,5120,0.013389333254761167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,2048,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,2048,0.013784888717863293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,1024,0.0038906666967603895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,1536,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,1536,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,3584,0.008330666356616551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,1024,0.00630044440428416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,768,0.0035413333939181436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,1536,0.013045333325862885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,3072,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,1024,0.011065777805116443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,512,0.003142222141226133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,768,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,768,0.0063164445261160536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,2560,0.008978666530715095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,256,0.0027795556104845474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,512,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,512,0.005893333504597346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,1536,0.004784000002675586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,256,0.0058755555914507965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,128,0.0025048889219760895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,256,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,256,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,128,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,64,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,128,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,32,0.002521777732504739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,65536,0.07919110854466756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,128,0.007637333538797166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,1024,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,64,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,32,0.005542222410440445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,51200,0.059395558304256864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,65536,0.06217510832680596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,65536,0.10845066441429986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,16384,0.026679111851586237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,51200,0.08535110950469971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,768,0.011053333679835001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,51200,0.049633777803844877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,65536,0.15371199448903403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,12288,0.021202666891945735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,16384,0.01956533392270406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,51200,0.12257244851854111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,16384,0.030237333642111883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,10240,0.01884533299340142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,12288,0.016137777103318107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,12288,0.02437688906987508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,8192,0.014025777578353882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,10240,0.01420177850458357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,16384,0.04663377669122484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,7168,0.012666666673289405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,12288,0.03755377729733785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,10240,0.021475555168257818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,8192,0.011533333195580376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,8192,0.018453333112928603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,6144,0.011325333681371478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,10240,0.03326311045222812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,7168,0.010972444381978778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,5120,0.009986666341622671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,8192,0.02844088938501146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,6144,0.010239111052619087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,6144,0.014898666077189975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,7168,0.025895110434956018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,4096,0.008597333398130205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,5120,0.009373333719041612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,5120,0.01240888900227017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,6144,0.024111110303137038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,3584,0.00793511089351442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,4096,0.00869599978129069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,4096,0.0110880004035102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,5120,0.021340444684028625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,3072,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,3584,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,512,0.010660444696744284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,4096,0.018220444520314533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,2560,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,3072,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,3584,0.017317333155208163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,3072,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,2048,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,2560,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,2560,0.008662222160233391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,3072,0.015856888559129503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,1536,0.0046693334976832075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,2048,0.007341333561473423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,2048,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,7168,0.016747555798954435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,1024,0.0038986665507157645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,2048,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,1536,0.0069102222720781965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,768,0.0035582222044467926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,1536,0.01276088919904497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,1024,0.006298666612969504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,1024,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,512,0.003157333367400699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,768,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,768,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,3584,0.010311110979980892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,512,0.005633777628342311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,512,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,256,0.0027351110345787476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,768,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,128,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,256,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,512,0.01000088867213991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,256,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,64,0.002193777718477779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,128,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,128,0.0058853332367208265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,256,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,2560,0.015136889285511441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,32,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,65536,0.05642933315700955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,64,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,1536,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,128,0.008006222546100616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,51200,0.049011554982927114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,32,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,65536,0.06878755489985149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,65536,0.06093688805898031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,16384,0.020968000094095867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,51200,0.0543813341193729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,51200,0.04868888854980469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,1024,0.011376889215575324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,12288,0.019204444355434842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,16384,0.019296000401178997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,65536,0.15395555231306288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,10240,0.01573244399494595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,51200,0.12202400631374782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,12288,0.01704266667366028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,12288,0.015786666009161208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,8192,0.013345777988433838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,12288,0.037112000915739275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,10240,0.013552889227867126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,10240,0.015186667442321777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,7168,0.01200888885392083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,8192,0.01313866674900055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,8192,0.011679111255539788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,10240,0.03268444538116455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,6144,0.010713777608341642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,7168,0.012116444607575735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,8192,0.027994665834638808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,6144,0.011092444260915121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,5120,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,6144,0.010041777458455827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,5120,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,6144,0.023522666758961145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,16384,0.020671111014154222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,5120,0.00980799976322386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,4096,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,3584,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,4096,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,5120,0.021105777886178758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,16384,0.04610310991605123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,3072,0.006800888727108638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,4096,0.008593777815500895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,3584,0.008041777544551426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,3584,0.00831911133395301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,2560,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,4096,0.01790844400723775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,3072,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,3072,0.0076248885856734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,2048,0.004868444469239977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,2560,0.007054222126801808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,7168,0.01072177787621816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,3072,0.015806222955385845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,1536,0.0041991112132867174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,2048,0.007140444384680853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,2560,0.01426755554146237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,2048,0.006858666737874349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,1024,0.003495111233658261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,7168,0.025928000609079998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,1536,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,1536,0.0064239998658498125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,768,0.00315022220214208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,2048,0.013459555804729462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,1024,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,1024,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,512,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,1536,0.012473777764373355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,768,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,768,0.0107004443804423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,512,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,256,0.0025066667132907445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,3584,0.016914667354689706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,512,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,128,0.002463999928699599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,256,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,256,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,256,0.009752888646390703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,128,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,64,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,32,0.002465777720014254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,128,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,65536,0.04875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,128,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,64,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,32,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,51200,0.03996710975964864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,768,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,65536,0.059453335073259145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,65536,0.06618311007817586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,16384,0.016960889101028442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,512,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,51200,0.04734399914741516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,51200,0.05219022101826138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,65536,0.1529137823316786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,12288,0.014084445105658637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,16384,0.020392000675201416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,2560,0.0074444446298811175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,51200,0.12150043911404079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,10240,0.013352000051074557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,12288,0.016660445266299777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,16384,0.04582488867971632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,10240,0.01308444473478529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,10240,0.014771555860837301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,8192,0.01238222254647149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,12288,0.03692177931467692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,10240,0.0324248903327518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,8192,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,7168,0.011992000043392181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,1024,0.011614222493436603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,8192,0.012288889123333825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,6144,0.010590222146775989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,7168,0.010703110860453712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,7168,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,6144,0.010628444453080496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,5120,0.00925422211488088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,7168,0.025626666016048853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,6144,0.00997688869635264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,16384,0.018590221802393597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,4096,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,5120,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,5120,0.009416888985368941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,4096,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,4096,0.008549333446555668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,5120,0.019908444748984445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,3584,0.008299555215570662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,3584,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,3072,0.00648266656531228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,4096,0.017765333255132038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,3072,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,3584,0.01640888883007897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,2560,0.0057608890864584185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,8192,0.02756799923049079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,3072,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,2048,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,2560,0.007621333003044128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,3072,0.015636444091796875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,2560,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,1536,0.00434755575325754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,2048,0.00666844430896971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,2048,0.007174222005738153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,2560,0.01455200049612257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,6144,0.02258488867017958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,1024,0.0036231109665499795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,1536,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,1536,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,3584,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,768,0.0032844444115956626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,1024,0.006540444576077991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,1024,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,1536,0.012320000264379712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,768,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,768,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,1024,0.011006222003036074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,256,0.002513777878549364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,768,0.01038933296998342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,512,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,512,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,128,0.002516444358560774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,256,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,256,0.00519822206762102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,512,0.010013333625263637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,64,0.0022479999396536085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,128,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,128,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,256,0.009297777381208207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,65536,0.03689600030581156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,12288,0.015009777413474189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,64,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,128,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,51200,0.029343111647499934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,32,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,2048,0.013578666581047906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,65536,0.05783910883797539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,65536,0.06361511018541124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,16384,0.014076444837782117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,512,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,51200,0.04670400089687771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,12288,0.011759999725553723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,51200,0.05024977856212192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,16384,0.016706667012638517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,65536,0.152636448542277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,16384,0.01994311147265964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,10240,0.010732444624106089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,12288,0.014131555954615274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,12288,0.016048888365427654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,16384,0.04554311103290982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,8192,0.012509333590666452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,10240,0.012806221842765808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,12288,0.03643911083539327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,7168,0.011241777903503843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,32,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,8192,0.012271111210187277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,8192,0.011387555963463254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,6144,0.01018488903840383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,7168,0.010682666467295753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,8192,0.026714666022194758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,7168,0.011349333657158745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,6144,0.010398222340477837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,6144,0.0099982221921285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,7168,0.024821332759327356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,6144,0.02253066665596432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,5120,0.009124444590674506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,4096,0.0075733330514695905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,51200,0.12096267276340061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,5120,0.019833778341611225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,4096,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,4096,0.008636444807052612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,3584,0.0068400001360310455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,10240,0.03212889035542806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,4096,0.017655111021465726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,3584,0.007672000262472365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,3072,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,3584,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,3584,0.016552888684802584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,3072,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,2560,0.005207111024194294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,5120,0.008802666432327693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,3072,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,2048,0.004570666700601578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,2560,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,3072,0.015488000379668342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,2560,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,1536,0.003903111235962974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,2048,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,2048,0.007213333414660559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,2560,0.014158222410413953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,5120,0.009184000392754873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,1024,0.0031742221779293488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,1536,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,1536,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,10240,0.013819555441538492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,2048,0.013431111143694984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,1024,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,1024,0.006335111127959357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,512,0.00254488881263468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,1536,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,768,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,768,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,256,0.002376000086466471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,512,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,768,0.011067555182509951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,512,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,128,0.0021235555824306277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,512,0.009943111075295342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,256,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,64,0.0019075555933846368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,128,0.004952888935804367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,128,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,32,0.0021582223061058256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,256,0.009708444277445475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,65536,0.030068443881140813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,128,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,32,0.004962666581074397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,51200,0.024691555235120986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,768,0.0028888889484935333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,65536,0.05751733647452461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,65536,0.06221244732538859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,16384,0.011502222054534487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,51200,0.04603022336959839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,1024,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,65536,0.15198577774895564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,51200,0.04899466700024075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,12288,0.011328889264000786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,16384,0.019273777802785236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,16384,0.016872000363137987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,10240,0.010319111247857412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,51200,0.12063644991980658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,12288,0.014097778333557976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,256,0.005647999958859549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,8192,0.00925333301226298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,16384,0.045447998576694064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,10240,0.012793777717484368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,12288,0.0355404449833764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,10240,0.013792888985739814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,7168,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,8192,0.011357333097192975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,8192,0.011932444241311816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,64,0.005175999883148405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,6144,0.008079110748238033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,7168,0.011247999966144562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,7168,0.010681777364677854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,5120,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,8192,0.02639644510216183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,6144,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,5120,0.009324444664849175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,5120,0.009164444274372524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,6144,0.02222577730814616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,4096,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,5120,0.019871110717455547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,3584,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,4096,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,12288,0.01514844430817498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,4096,0.017648888958825003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,3584,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,3072,0.0057502223385704895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,3584,0.007840000092983246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,10240,0.031424889961878456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,3584,0.016539555456903245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,2560,0.005136888888147143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,3072,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,6144,0.00999466660949919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,3072,0.015513777732849121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,2048,0.00452533322903845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,2560,0.007045333584149678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,2560,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,2560,0.014484445254007975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,2048,0.0069884442620807225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,4096,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,1536,0.004138666722509596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,2048,0.013400889105266996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,1536,0.006679111056857639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,1024,0.003294222263826264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,1536,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,1024,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,1536,0.01205244412024816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,1024,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,768,0.0031537777847713898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,3072,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,512,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,768,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,1024,0.01108888867828581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,768,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,7168,0.024100444383091394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,256,0.002483555633160803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,512,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,512,0.005380444642570283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,768,0.01015022231472863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,128,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,256,0.0058453331391016645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,256,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,64,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,128,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,128,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,768,32,0.0021155555215146807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,65536,0.023753777146339417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,128,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,64,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,32,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,51200,0.018752889500723947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,65536,0.06100800302293566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,65536,0.05703910854127672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,51200,0.04818755719396803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,65536,0.15161866611904568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,16384,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,16384,0.016872000363137987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,51200,0.12088533242543538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,12288,0.01237511138121287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,16384,0.018569777409235638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,512,0.010037333601050908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,10240,0.010984888507260216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,2048,0.006744888921578725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,16384,0.04520533482233683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,12288,0.013969777358902825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,8192,0.009764444496896531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,10240,0.012770666844315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,10240,0.013518222504191928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,12288,0.03557066784964667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,7168,0.008656889200210571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,8192,0.011767999993430244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,8192,0.011335999601417117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,10240,0.03116977877087063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,6144,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,7168,0.010715555813577441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,7168,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,8192,0.026100445124838088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,5120,0.007370666497283512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,6144,0.01034044474363327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,6144,0.01000088867213991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,5120,0.009354666703277165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,5120,0.009190222455395592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,6144,0.021944888763957556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,4096,0.006671999891599019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,256,0.009133333133326637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,4096,0.008548444343937768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,3584,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,12288,0.015136889285511441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,4096,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,3072,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,3584,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,4096,0.01755288905567593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,3584,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,2560,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,3072,0.0075288886825243635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,3072,0.007700444095664554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,3584,0.016128000285890367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,2048,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,2560,0.0075866662793689305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,51200,0.04486844605869717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,2560,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,3072,0.015456888410780164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,1536,0.0038862224254343244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,7168,0.02404444416364034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,2048,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,1024,0.003176888864901331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,1536,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,1536,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,2048,0.013088000317414602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,1536,0.012434666355450949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,1024,0.005744888964626525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,768,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,5120,0.019634667370054457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,1024,0.006455111006895701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,512,0.002524444419476721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,768,0.006222222414281633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,1024,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,768,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,256,0.0022088889446523455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,512,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,512,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,768,0.011048888994587792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,128,0.002199999988079071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,256,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,256,0.005115555392371284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,512,0.010018666585286459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,64,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,128,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,128,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,2048,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,256,0.009903111391597325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,2560,0.014123555686738757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,65536,0.0169840008020401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,128,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,64,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,65536,0.053970668050977916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,51200,0.014621333943472968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,65536,0.05922044648064507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,16384,0.009639999932712978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,51200,0.040376888381110296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,51200,0.04760266674889458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,65536,0.15124889214833578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,12288,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,16384,0.018215111560291715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,16384,0.016522667474216886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,51200,0.11997955375247532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,10240,0.007668444679843054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,12288,0.013824000126785703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,16384,0.044023109806908496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,10240,0.01277955538696713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,12288,0.035078220897250705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,512,32,0.0019377778387731975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,8192,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,8192,0.011352000137170156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,7168,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,10240,0.030822220775816176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,32,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,7168,0.01032977799574534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,6144,0.0063075555695427795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,7168,0.011018666956159802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,6144,0.009838222629494136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,5120,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,6144,0.010012444522645738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,12288,0.014816888504558139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,6144,0.02165511084927453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,4096,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,5120,0.008973333570692275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,5120,0.009045333498054082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,5120,0.01956977778010898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,4096,0.008360889222886827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,4096,0.008086221913496653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,8192,0.011699555648697747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,4096,0.017619555195172627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,3584,0.007695111135641734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,3072,0.005444444302055571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,8192,0.02644799980852339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,3584,0.016177778442700703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,3072,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,2560,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,7168,0.023804444405767653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,3072,0.015508444772826301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,2560,0.0069919998447100324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,2048,0.004214222232500712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,2560,0.014123555686738757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,10240,0.013355555633703867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,1536,0.003682666768630346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,2048,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,2048,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,3584,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,1024,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,2048,0.013457777599493662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,1536,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,1536,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,768,0.0028320000403457214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,1024,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,3584,0.008094222181373173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,1024,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,1536,0.012411555482281579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,512,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,768,0.006217777729034424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,768,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,1024,0.010857777463065254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,3072,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,512,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,768,0.010732444624106089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,2560,0.007328888608349695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,128,0.0018586667461527719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,256,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,512,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,256,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,64,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,256,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,32,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,128,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,65536,0.014735110931926303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,64,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,128,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,51200,0.012466666599114736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,32,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,65536,0.049377779165903725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,65536,0.058724443117777504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,16384,0.00703999979628457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,51200,0.040044443474875555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,65536,0.15085511737399632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,256,256,0.002181333386235767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,12288,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,16384,0.018231110440360177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,16384,0.016529776983790927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,512,0.005306666509972678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,51200,0.1192951069937812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,10240,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,12288,0.013818666338920593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,16384,0.04413333204057482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,8192,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,10240,0.013176889055305056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,10240,0.012533333566453723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,12288,0.034910221894582115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,8192,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,7168,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,10240,0.03091111116939121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,8192,0.011496889094511667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,6144,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,7168,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,8192,0.025921778546439275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,5120,0.006315555423498154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,6144,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,6144,0.009995555712117089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,7168,0.023885332875781592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,51200,0.04586133360862732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,4096,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,5120,0.00903733323017756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,5120,0.008754666480753157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,6144,0.02200355629126231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,3584,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,4096,0.008416888614495596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,4096,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,5120,0.019595555133289762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,3072,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,12288,0.014733332726690503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,4096,0.01723377737734053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,3584,0.008084444536103142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,3584,0.007680888805124495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,2560,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,3072,0.007677333222495184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,3584,0.016217778126398723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,2048,0.004600888739029567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,3072,0.007068444457319047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,2560,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,1536,0.003680888977315691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,2560,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,3072,0.015207111835479736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,2048,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,1024,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,2560,0.014484445254007975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,2048,0.006330666856633292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,1536,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,7168,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,1536,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,768,0.0028080000645584534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,2048,0.012740444805887012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,1024,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,512,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,1024,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,768,0.006096000058783426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,1536,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,256,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,768,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,512,0.005667555663320754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,1024,0.011384000380833944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,128,0.0019306666735145782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,512,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,768,0.01033333357837465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,256,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,256,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,128,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,32,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,512,0.009687999884287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,65536,0.011295110815101199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,128,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,51200,0.010006222460005019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,16384,0.005889777921968036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,64,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,128,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,32,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,10240,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,65536,0.05646222167544895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,51200,0.04553688897026909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,16384,0.01795111099878947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,8192,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,7168,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,12288,0.014855111638704935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,6144,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,10240,0.01312444441848331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,5120,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,8192,0.011583999627166323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,4096,0.005335999859703912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,6144,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,3584,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,5120,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,3072,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,4096,0.007991111112965478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,128,64,0.0017893332988023758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,2560,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,2048,0.004848888764778773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,3584,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,256,0.009747555686367882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,3072,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,1024,0.002874666617976295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,2560,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,768,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,12288,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,2048,0.00657066661450598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,512,0.0024622221373849446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,256,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,1536,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,128,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,1024,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,64,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,768,0.005281777845488654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,512,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,32,0.001806222316291597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,256,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,128,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,51200,0.013747555514176687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,7168,0.010967999696731567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,64,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,16384,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,32,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,12288,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,10240,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,65536,0.05647289090686374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,8192,0.005664888769388199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,51200,0.04520533482233683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,16384,0.017903111047214933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,7168,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,12288,0.014734221829308404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,6144,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,5120,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,10240,0.01333333303531011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,64,1536,0.003552000141806073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,4096,0.004560889055331548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,3584,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,8192,0.011418666276666852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,7168,0.010767111347781287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,3072,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,2560,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,5120,0.008839999636014303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,4096,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,2048,0.004544888933499654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,1536,0.0035413333939181436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,3584,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,1024,0.002887111157178879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,3072,0.007242666350470648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,768,0.0026666666898462507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,512,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,2560,0.00665511108107037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,256,0.0021964444054497615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,65536,0.016192000773217943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,2048,0.006360888895061281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,128,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,1536,0.006031111296680238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,64,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,1024,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,256,32,32,0.001810666587617662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,768,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,512,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,256,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,128,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,64,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,32,0.004854222138722737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,12288,0.3200017876095242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,16384,0.4911857710944281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,6144,0.010000000397364298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,16384,0.3431386682722304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,16384,0.14168977737426758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,10240,0.2464240127139621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,12288,0.2645920117696126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,10240,0.2142782211303711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,10240,0.09070666631062825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,51200,1.0962488386366103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,12288,0.10748800304200913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,7168,0.18107199668884277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,51200,0.4226586553785536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,8192,0.17635733551449248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,8192,0.07338044378492567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,6144,0.15187466144561768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,5120,0.1306942171520657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,6144,0.05743110842174954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,6144,0.14211022853851318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,5120,0.11061955822838677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,5120,0.0487973325782352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,8192,0.20264355341593424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,4096,0.041166222757763334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,4096,0.09140800105200873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,3584,0.09221866395738389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,51200,1.5231982337103949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,3584,0.03633244501219855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,3072,0.08106666803359985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,7168,0.06557599703470866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,3584,0.0793271130985684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,3072,0.03286666671435038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,2560,0.028530667225519817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,2560,0.06005599763658312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,2048,0.05507022142410278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,3072,0.06858755482567681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,1536,0.04279555545912849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,4096,0.1134808858235677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,1024,0.032568001084857516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,2048,0.025062221619817946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,2048,0.05055377880732218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,1536,0.019648000597953796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,768,0.0255724440018336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,1536,0.03987911012437608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,1024,0.015775110986497667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,512,0.017952000101407368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,1024,0.031609776947233416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,256,0.011401777466138205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,512,0.013155555559529198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,512,0.02144355575243632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,7168,0.1528666681713528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,256,0.01885511146651374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,256,0.011920000116030375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,2560,0.0672631131278144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,64,0.007927999728255803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,32,0.008266666697131263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,128,0.011730666789743634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,128,0.018378666705555387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,64,0.017892445127169292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,768,0.014208000567224292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,768,0.02545333405335744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,128,0.010088000032636855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,65536,1.06550841861301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,51200,1.0305911170111763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,65536,1.3165039486355252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,16384,0.32776978280809194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,32,0.017902221944597032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,12288,0.2589671081966824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,51200,0.8566853205362955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,12288,0.08451288938522339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,12288,0.19997155666351318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,16384,0.1129937834209866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,10240,0.17435199684566924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,8192,0.1446604463789198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,51200,0.33094578319125706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,10240,0.1746000051498413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,7168,0.12535999880896673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,8192,0.05896266963746813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,6144,0.10885333352618748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,7168,0.05220266514354282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,65536,0.42058666547139484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,5120,0.09294578101899888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,7168,0.1332133346133762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,16384,0.2689431243472629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,4096,0.07850044303470187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,6144,0.10535999801423813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,6144,0.045765333705478244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,5120,0.03913777735498216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,4096,0.03403644429312812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,5120,0.08915822373496161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,4096,0.07157422436608209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,3584,0.029976887835396662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,3584,0.06320799721611871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,2560,0.048998223410712353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,10240,0.07112266620000203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,2048,0.044547554519441396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,3072,0.05526399943563673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,2560,0.0240639994541804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,3072,0.026878222823143005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,2560,0.047911110851499766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,1536,0.031360887818866305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,8192,0.1450257831149631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,1024,0.022709333234363135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,2048,0.020284444093704224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,2048,0.041160000695122614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,768,0.017659554878870647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,1536,0.016599110431141324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,512,0.01372977760103014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,1536,0.03286044465170966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,1024,0.025604445073339675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,256,0.01092977739042706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,768,0.02093244426780277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,512,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,3584,0.06614666514926486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,768,0.012780444489585029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,512,0.017733333839310538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,256,0.01034400032626258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,256,0.016027554869651794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,128,0.010407110883129967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,3072,0.057886222998301186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,32,0.00829777783817715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,128,0.015136889285511441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,64,0.015124445160230001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,32,0.014988443917698331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,65536,0.3881804413265652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,51200,0.3056088818444146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,1024,0.013896889156765409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,65536,0.3614071210225423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,65536,0.1466577715343899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,16384,0.10861866341696845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,51200,0.2858533329433865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,12288,0.0825493335723877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,128,0.008990222381220924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,64,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,16384,0.04140799906518724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,10240,0.07332089212205675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,16384,0.09258666965696548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,51200,0.1142595609029134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,12288,0.07012089093526204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,12288,0.0324888891643948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,10240,0.027994665834638808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,7168,0.05060888992415535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,8192,0.02379911144574483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,8192,0.04944088723924425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,6144,0.04416888952255249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,7168,0.02127644419670105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,5120,0.03741688860787286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,7168,0.04371466570430332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,6144,0.019090667366981506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,4096,0.031186666753556993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,6144,0.0384808878103892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,5120,0.01687466601530711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,3584,0.028188443846172754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,5120,0.032920890384250216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,4096,0.015049777097172208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,3072,0.02496622171666887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,4096,0.02791822287771437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,3584,0.01385155568520228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,2560,0.020611556039916146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,3584,0.02495555579662323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,3072,0.01291822228166792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,2048,0.01683911184469859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,3072,0.022065778573354084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,2560,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,2048,0.010905777414639791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,2560,0.019433778193261888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,8192,0.05945599741405911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,1024,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,2048,0.01700533264213138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,768,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,1536,0.014174222946166992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,10240,0.05908888578414917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,1024,0.007821333077218797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,1024,0.011801777614487542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,768,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,256,0.004668444395065308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,768,0.01001422190003925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,512,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,128,0.003908444609906939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,256,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,512,0.008836444053384993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,64,0.003946666502290302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,128,0.00676711110605134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,32,0.004204444587230682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,128,0.008330666356616551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,64,0.008294222255547842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,32,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,65536,0.3030480013953315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,1536,0.013382222089502545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,51200,0.23500800132751465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,65536,0.29321334097120494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,16384,0.07625510957505968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,1536,0.009038222332795462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,65536,0.1283982197443644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,51200,0.23052801026238334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,512,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,12288,0.059530668788486056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,10240,0.04986133178075155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,51200,0.1020088858074612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,16384,0.07684444718890719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,16384,0.0377768874168396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,8192,0.044237332211600415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,12288,0.030279109875361126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,7168,0.0370053317811754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,12288,0.058548443847232394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,10240,0.050208889775806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,256,0.008367111285527548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,8192,0.021951110826598272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,8192,0.04215911030769348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,5120,0.028985778490702312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,7168,0.01992266707950168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,7168,0.037272890408833824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,4096,0.023406222462654114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,6144,0.01807466646035512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,5120,0.016140444411171805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,3584,0.020576000213623047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,5120,0.02786311176088121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,4096,0.014507555299335055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,6144,0.0329751107427809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,3072,0.018185777796639335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,3584,0.01350488927629259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,4096,0.02348800003528595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,2560,0.015209777487648858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,3584,0.021378666162490845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,3072,0.012617777619096967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,3072,0.01923199991385142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,2560,0.01720266706413693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,1536,0.010044444766309526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,2048,0.010072888599501716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,2048,0.014813333749771118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,1024,0.008657777474986183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,1536,0.008345777789751688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,1536,0.012694222231705984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,10240,0.02569955587387085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,6144,0.03220622075928582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,768,0.007348444726732042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,512,0.005663110978073544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,1024,0.007663999994595845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,1024,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,256,0.004536889079544279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,768,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,512,0.007093333535724216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,768,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,128,0.003543111185232798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,512,0.008036444584528605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,256,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,64,0.0031973332580592898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,256,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,32,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,128,0.006771555377377405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,128,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,2048,0.012995555169052549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,64,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,65536,0.23844178517659506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,32,0.007618666523032718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,16384,0.0610426664352417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,51200,0.18631554974450004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,65536,0.24775645467970106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,12288,0.048173334863450795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,16384,0.03638666537072923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,51200,0.09910577535629272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,16384,0.06582133637534247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,10240,0.04500088757938809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,8192,0.03378933336999681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,12288,0.028504000769721136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,10240,0.024772443705134924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,12288,0.050999999046325684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,7168,0.029578665892283123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,10240,0.043362667163213096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,8192,0.021213332811991375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,6144,0.026593777868482802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,8192,0.03639466563860575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,5120,0.023194667365815904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,7168,0.03200177682770623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,6144,0.01752800080511305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,4096,0.01928444372283088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,6144,0.02829155657026503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,2560,0.011696889168686338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,5120,0.01572800013754103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,3584,0.016719111137919955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,5120,0.024456888437271118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,3072,0.014945778581831189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,4096,0.014164444473054675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,4096,0.02068355513943566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,65536,0.12496266100141738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,2560,0.012755555411179861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,3584,0.018784888916545443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,3072,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,2048,0.010625777973069085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,3072,0.01718311177359687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,2560,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,1536,0.008981333010726506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,2560,0.015219555960761176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,2048,0.008977777428097194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,1024,0.007239999870459239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,1536,0.008059555457697975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,2048,0.013438222308953604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,768,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,1024,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,1024,0.009138666921191746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,512,0.004621333132187526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,7168,0.019312888383865356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,768,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,256,0.0038542221817705366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,512,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,128,0.0033235556135574975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,512,0.007510222494602203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,256,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,64,0.0031911111954185697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,256,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,128,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,32,0.003294222263826264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,128,0.006677333265542984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,3584,0.01312622262371911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,64,0.0068897778789202375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,51200,0.1928373310301039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,65536,0.24215822749667695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,51200,0.18659822146097818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,65536,0.07671022415161133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,65536,0.20309778054555258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,16384,0.06675555308659871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,1536,0.011392888923486074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,51200,0.15713955296410456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,12288,0.0510791109667884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,51200,0.06231377522150675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,16384,0.054698665936787925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,768,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,10240,0.04295733239915636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,8192,0.03459466828240289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,12288,0.01925155520439148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,10240,0.016890666551060148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,12288,0.04229955540763008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,7168,0.030943112240897283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,10240,0.036095112562179565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,8192,0.015225778023401896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,6144,0.02647022240691715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,8192,0.0302800006336636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,7168,0.01387288918097814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,5120,0.022679110368092854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,7168,0.026770666241645813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,32,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,6144,0.023706666297382776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,4096,0.018736888964970905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,5120,0.011696889168686338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,3584,0.016355555918481614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,5120,0.02067377832200792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,3072,0.01462044484085507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,3584,0.009757333331637913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,4096,0.017826666434605915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,3584,0.016342222690582275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,2560,0.012597333225939008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,3072,0.008778666456540426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,3072,0.015001777145597668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,16384,0.02386577758524153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,2048,0.0107004443804423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,2560,0.008371555142932469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,1536,0.008896888958083259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,2560,0.013447999954223633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,2048,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,1024,0.007127111156781514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,1536,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,2048,0.012076444096035428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,768,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,1536,0.010043555663691627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,512,0.0041848888827694785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,768,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,1024,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,256,0.003506666670242945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,768,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,512,0.006228444476922353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,512,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,128,0.0031537777847713898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,256,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,64,0.002883555574549569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,256,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,4096,0.010833777487277985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,128,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,128,0.006319111006127463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,64,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,32,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,51200,0.15442311763763428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,65536,0.19821600119272867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,65536,0.0728400018480089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,16384,0.05263466636339823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,1024,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,12288,0.0376160012351142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,51200,0.059021333853403725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,16384,0.023006222314304773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,10240,0.034097777472601995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,16384,0.05314044488800896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,6144,0.01278222186697854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,12288,0.018527110417683918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,8192,0.027038221557935078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,12288,0.04143822193145752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,10240,0.01644355555375417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,7168,0.023341332872708637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,32,0.0031537777847713898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,10240,0.03524355424775018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,6144,0.0214382227924135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,8192,0.029618667231665716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,5120,0.018607111440764535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,65536,0.17934400505489775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,7168,0.026258667310078938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,4096,0.015105777316623263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,6144,0.023367111881573994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,5120,0.011377777490350934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,5120,0.02017244365480211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,3584,0.013594666288958656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,4096,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,3072,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,4096,0.01745244363943736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,3584,0.009335111412737105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,2560,0.011017777853541903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,3584,0.01610133383009169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,3072,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,2048,0.008978666530715095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,51200,0.1538542244169447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,3072,0.014848889576064216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,2560,0.00777599960565567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,1536,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,2560,0.013129777378506131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,2048,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,1024,0.006340444501903322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,2048,0.011335111326641507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,1536,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,1536,0.009483555952707926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,768,0.005537777725193236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,1024,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,7168,0.013444444371594323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,512,0.004595555365085602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,1024,0.008084444536103142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,6144,0.012459555433856117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,256,0.003552888830502828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,768,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,128,0.0031511110977994073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,512,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,256,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,64,0.0028115556471877625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,256,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,128,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,32,0.0028400001012616684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,128,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,64,0.00630044440428416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,65536,0.16762222184075248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,32,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,51200,0.13139822747972277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,65536,0.1597493357128567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,65536,0.0710044436984592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,16384,0.04606399933497111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,8192,0.014481777946154276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,51200,0.12373333507113987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,12288,0.036120888259675764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,51200,0.057320886188083224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,16384,0.021838222940762837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,10240,0.030756443738937378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,16384,0.04389422138532003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,8192,0.025553777813911438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,512,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,12288,0.01793955597612593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,7168,0.021960000197092693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,10240,0.015986666083335876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,8192,0.014041778114106921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,10240,0.028913779391182795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,6144,0.019278221660190158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,7168,0.013219555219014486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,8192,0.02460977766248915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,5120,0.016912000046836007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,7168,0.021945777866575453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,6144,0.012048888537618848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,4096,0.013553777502642738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,6144,0.0196079992585712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,3584,0.01262399968173769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,5120,0.017252445220947266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,4096,0.0094355551732911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,3072,0.011136000355084738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,4096,0.015056000815497505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,3584,0.00903200027015474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,2560,0.010321777727868822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,3072,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,2048,0.008309333688682979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,2560,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,3072,0.012831110921170978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,1536,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,2560,0.011511999699804517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,768,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,2048,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,1024,0.005203555441564984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,2048,0.009944000177913243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,12288,0.03432710965474447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,768,0.0042453333735466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,1536,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,512,0.0035911111368073356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,1024,0.006562666760550604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,768,0.00630666646692488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,1024,0.007328888608349695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,256,0.0028799999919202593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,768,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,512,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,128,0.0026400000270869997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,512,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,256,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,5120,0.011001778145631155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,32,0.002516444358560774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,128,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,256,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,128,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,65536,0.1456071138381958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,64,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,32,0.006018666757477655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,3584,0.013839999834696451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,51200,0.11400532722473145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,65536,0.06799911128150092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,65536,0.15359111626942953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,12288,0.03229511115286086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,51200,0.055000000529819064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,16384,0.021300445000330608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,10240,0.028063999281989202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,16384,0.042415999703937106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,12288,0.01745244363943736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,1536,0.008575110799736446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,12288,0.03292177783118354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,8192,0.023100444012218054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,10240,0.015454222758611044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,7168,0.02021511064635383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,10240,0.02794577678044637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,8192,0.023903111616770428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,64,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,6144,0.018241778016090393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,7168,0.01274577776590983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,5120,0.014902222487661572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,7168,0.02132622235351139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,6144,0.011769777370823754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,4096,0.013771555489963956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,6144,0.01935822268327077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,5120,0.010618666807810465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,3584,0.012039110892348819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,16384,0.04271999994913737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,4096,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,5120,0.016680000556839835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,3072,0.010021333065297868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,4096,0.01482399966981676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,2560,0.009927999642160203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,51200,0.11975110901726617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,3072,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,3584,0.013756444056828817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,2048,0.00811822215716044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,3072,0.012497777740160624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,2560,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,1536,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,2560,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,2048,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,1024,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,1536,0.0069120000633928515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,768,0.004232888834344016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,8192,0.013566222455766467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,1024,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,512,0.0036151111125946045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,1024,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,768,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,256,0.0028817777832349143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,512,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,512,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,128,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,256,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,256,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,64,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,128,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,3584,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,32,0.002806222273243798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,65536,0.13103289074367946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,64,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,32,0.006088888893524806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,51200,0.0959377752410041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,2048,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,65536,0.0654675563176473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,16384,0.038796444733937584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,1536,0.008296888735559251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,65536,0.11666844950781928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,12288,0.029655112160576716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,768,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,51200,0.052461332745022245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,10240,0.024823110964563157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,16384,0.020228443874253165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,8192,0.020269332660569083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,12288,0.016894222961531747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,16384,0.03169511093033685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,12288,0.02551199992497762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,7168,0.017854221993022494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,10240,0.01514400045077006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,6144,0.015650666422314115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,8192,0.013432888521088494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,10240,0.02234933276971181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,7168,0.012296000288592445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,5120,0.013448889056841532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,8192,0.01904444396495819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,4096,0.01186222251918581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,7168,0.017095999585257638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,6144,0.010849778023031024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,128,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,3584,0.010101333260536194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,5120,0.009720000127951304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,6144,0.015847999188635085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,3072,0.009928888744778102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,4096,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,5120,0.013787555197874704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,2560,0.00850133349498113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,4096,0.012264000044928657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,3072,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,2048,0.007155555817815993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,3584,0.011122666299343109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,1536,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,2560,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,2560,0.008939555949634975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,2048,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,1024,0.004545777622196409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,1536,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,768,0.0041831110914548235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,1536,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,1024,0.006462222172154321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,51200,0.08918222453859116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,512,0.003531555748648114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,1024,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,768,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,256,0.0028693332440323303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,768,0.006302222195598815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,512,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,128,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,512,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,256,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,64,0.002524444419476721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,256,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,128,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,32,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,128,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,65536,0.10749600330988567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,64,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,3072,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,32,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,51200,0.08287377489937676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,16384,0.03216800093650818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,65536,0.06433422035641141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,65536,0.11449066797892253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,2048,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,12288,0.027792887555228338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,51200,0.05130133363935682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,16384,0.019952888290087383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,10240,0.02328355610370636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,51200,0.08983644511964585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,16384,0.03181511163711548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,8192,0.019235556324323017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,12288,0.016548444827397663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,7168,0.016923555069499545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,10240,0.014937778313954672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,12288,0.025472000241279602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,6144,0.014808000789748298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,8192,0.01275288893116845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,5120,0.013048888908492194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,7168,0.011798222031858234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,6144,0.010599111517270407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,7168,0.017235555582576327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,4096,0.009735999835862054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,6144,0.015522667103343539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,5120,0.009369778136412302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,3584,0.009639110830095079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,5120,0.013800000150998434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,3584,0.008347555167145198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,3072,0.008591111335489485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,4096,0.011590222517649332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,3584,0.008290666672918532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,2560,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,3584,0.010728889041476779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,3072,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,2048,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,3072,0.009724444813198512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,1536,0.0058373332851462895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,2560,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,2560,0.00907199995385276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,2048,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,1024,0.0047608887155850725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,2048,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,1536,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,768,0.004212444441186057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,1024,0.006311999840868845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,1536,0.007300444775157505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,512,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,768,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,1024,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,10240,0.021926222576035395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,256,0.0027031112048361036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,512,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,128,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,8192,0.019155555301242404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,64,0.0021946666141351065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,256,0.005703111075692707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,32,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,256,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,4096,0.0086666668454806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,65536,0.08372977707121108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,32,0.005709333138333426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,51200,0.06775733497407702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,65536,0.10802844497892593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,16384,0.0285502208603753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,51200,0.05048977666431003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,51200,0.08528000116348267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,12288,0.025754666990704004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,16384,0.01977777812216017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,16384,0.030455999904208716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,10240,0.01800622211562263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,12288,0.016524444023768108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,12288,0.02455377744303809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,8192,0.014856000741322836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,512,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,128,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,10240,0.014656888114081489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,10240,0.021440888444582622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,7168,0.013284444808959961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,8192,0.012101333174440594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,6144,0.011805333197116852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,8192,0.018440888987647164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,64,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,7168,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,5120,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,7168,0.01686311099264357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,4096,0.009293333523803288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,5120,0.009413333402739631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,6144,0.015250666273964776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,65536,0.06326311164432101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,5120,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,4096,0.008648888932334052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,3072,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,3584,0.008285333712895712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,4096,0.011050666371981302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,2560,0.006793777975771163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,3584,0.010374222364690568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,3072,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,2048,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,3072,0.009704888694816166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,2560,0.0076248885856734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,1536,0.004607999904288186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,768,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,2560,0.008664888640244802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,1024,0.0038248888320393036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,2048,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,2048,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,768,0.003314666656984223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,1024,0.00647555540005366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,1024,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,512,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,768,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,768,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,6144,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,256,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,512,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,3584,0.008234666453467475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,128,0.0022515555222829184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,256,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,512,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,64,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,32,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,256,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,128,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,65536,0.08168888754314847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,128,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,64,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,32,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,65536,0.06186044216156006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,16384,0.02664711078008016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,65536,0.10638933049307929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,12288,0.020661332541041903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,51200,0.08368444442749023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,16384,0.019719110594855416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,1536,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,16384,0.029954665237002905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,10240,0.021770666042963665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,1536,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,12288,0.016146666473812528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,12288,0.023964444796244305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,10240,0.014262222581439547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,7168,0.012684444586435953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,10240,0.02113866640461816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,8192,0.011707555916574268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,6144,0.011159111228254108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,8192,0.018179555733998615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,7168,0.01071111112833023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,5120,0.009934222532643212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,7168,0.016422222057978313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,6144,0.01000533335738712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,4096,0.009119110802809397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,6144,0.014825777875052558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,5120,0.009364444348547194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,3584,0.008120888637171851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,5120,0.012424888710180918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,4096,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,3072,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,3584,0.008295111358165741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,51200,0.06571555799908109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,3072,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,3584,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,2048,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,3072,0.009392889009581672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,2560,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,1536,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,2560,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,2048,0.007198221981525421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,2048,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,1024,0.00435644429590967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,1536,0.006735111276308696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,768,0.0038693332009845306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,1024,0.0063946665161185795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,1536,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,512,0.003212444484233856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,1024,0.006548444430033366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,768,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,256,0.0028497777465316984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,768,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,512,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,512,0.00566133318675889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,128,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,256,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,256,0.005647111270162795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,4096,0.010687111152542962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,64,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,128,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,32,0.0022168887986077205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,128,0.0053137776752312975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,65536,0.069069332546658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,51200,0.04898666673236423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,64,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,32,0.00553955551650789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,51200,0.05262044403288099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,8192,0.014434667097197639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,16384,0.021920000513394673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,65536,0.06651822063657972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,12288,0.017937777770890128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,51200,0.048327111535602145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,51200,0.05275999837451511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,16384,0.018890667292806838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,10240,0.01739555597305298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,16384,0.020239111449983384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,12288,0.015385778413878547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,8192,0.013443555268976422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,12288,0.01683466633160909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,10240,0.01310222182008955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,7168,0.011959999799728394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,8192,0.011416888899273343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,10240,0.015032000011867948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,6144,0.010792888700962067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,8192,0.013105777402718862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,7168,0.010768888725174798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,5120,0.009633777870072259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,7168,0.01201777739657296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,6144,0.00998933364947637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,2560,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,4096,0.008282666405042013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,3584,0.0075493330756823225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,5120,0.009319999979601966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,5120,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,4096,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,3072,0.00686488880051507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,3584,0.008328888979223039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,4096,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,2560,0.0058240000572469496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,3584,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,3072,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,2560,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,3072,0.007596444752481248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,1536,0.00423555572827657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,2560,0.0070364442136552595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,2048,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,1024,0.0035422220826148987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,1536,0.006913777854707506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,1536,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,768,0.0031804444475306403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,1024,0.0063031112982167145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,1024,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,512,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,768,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,256,0.002456888970401552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,768,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,512,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,6144,0.010766222245163388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,512,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,256,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,256,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,128,0.005884444548024072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,32,0.002285333350300789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,128,0.005087999833954705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,2048,0.004961777892377642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,65536,0.04919999837875366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,32,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,2048,0.006689777804745569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,51200,0.039742221434911094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,65536,0.06418310933642916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,16384,0.016559999850061204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,51200,0.04672799838913811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,51200,0.050607111718919545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,12288,0.014552889598740472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,16384,0.01994311147265964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,16384,0.018360000517633226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,10240,0.015556444724400839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,128,0.002500444443689452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,12288,0.014133334159851074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,8192,0.013075555364290873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,12288,0.016460445192125108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,64,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,10240,0.012793777717484368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,7168,0.011762667033407422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,10240,0.014284445179833306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,8192,0.011370666325092316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,6144,0.010478222535716163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,65536,0.06044711007012261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,7168,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,5120,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,8192,0.012429333395428128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,64,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,4096,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,6144,0.00998133338159985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,7168,0.011228444675604502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,3584,0.007255111303594377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,6144,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,4096,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,5120,0.009160888691743216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,3072,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,4096,0.008303110798199972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,2560,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,3584,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,3584,0.008203555312421586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,3072,0.007691555553012424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,2048,0.005543111099137201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,2560,0.00739911115831799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,3072,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,1536,0.004801777915822135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,2560,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,2048,0.007204444044166141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,1024,0.0038542221817705366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,2048,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,1536,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,768,0.0035004446076022256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,1536,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,1024,0.006413333531883027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,512,0.0031786666562159858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,1024,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,768,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,256,0.0026195556339290407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,768,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,128,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,512,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,256,0.005191111316283544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,64,0.0021280000607172647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,5120,0.00925422211488088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,32,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,128,0.004977777600288391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,65536,0.05834666887919108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,65536,0.0365866654449039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,64,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,32,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,65536,0.057166225380367704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,65536,0.061498668458726674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,16384,0.014143110977278816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,51200,0.046239111158582896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,51200,0.048771556880739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,12288,0.012110222544935016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,16384,0.016785777277416654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,16384,0.019319999549123976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,10240,0.011005333728260465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,512,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,12288,0.014107555150985718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,256,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,8192,0.009931555224789513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,12288,0.015508444772826301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,7168,0.009964444571071202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,128,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,8192,0.011256888508796692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,10240,0.01367022262679206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,6144,0.00904622260067198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,8192,0.011747555600272285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,7168,0.010664888554149203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,7168,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,6144,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,4096,0.00741155528359943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,51200,0.029222223493787978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,6144,0.01016177816523446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,3584,0.006510222123728857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,5120,0.009008888569143083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,3072,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,4096,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,3584,0.00812000036239624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,2560,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,3584,0.007663999994595845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,3072,0.007879111501905654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,3072,0.007338666253619724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,2048,0.004551110996140374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,2560,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,1536,0.0038968887594011095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,2560,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,2048,0.007053333024183909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,1024,0.00318755561278926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,2048,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,1536,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,768,0.0028808888875775864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,1536,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,1024,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,512,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,1024,0.00564177789621883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,768,0.005984888722499211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,256,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,768,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,512,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,5120,0.00904444439543618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,512,0.0053191110491752625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,256,0.005767999837795894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,64,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,256,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,128,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,32,0.0021128888345426982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,65536,0.030413332912656996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,64,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,32,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,51200,0.024728000164031982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,65536,0.05712888638178507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,65536,0.06007733609941271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,16384,0.011921777493423887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,10240,0.01275288893116845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,51200,0.04587644338607788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,12288,0.011574221981896294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,51200,0.04747466577423943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,16384,0.016786666380034555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,5120,0.00871733327706655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,10240,0.010597333312034607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,16384,0.01886133352915446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,12288,0.014021333720948962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,8192,0.00959822204377916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,12288,0.015042665931913586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,10240,0.01274311128589842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,7168,0.008760000268618265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,4096,0.008389333056079017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,10240,0.01346577786737018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,8192,0.011095110740926532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,6144,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,5120,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,7168,0.010667555862002902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,8192,0.011741333537631564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,4096,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,6144,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,7168,0.01086400035354826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,3584,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,5120,0.009137777818573846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,4096,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,5120,0.009036444127559662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,3072,0.005668444352017508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,3584,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,2560,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,3584,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,3072,0.007663999994595845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,2048,0.004523555437723796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,3072,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,2560,0.007374222079912822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,1536,0.003907555507289039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,2560,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,2048,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,1024,0.0035128887328836652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,2048,0.006607111129495833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,1536,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,768,0.003128000017669466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,1536,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,1024,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,512,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,1024,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,768,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,256,0.0024968888610601425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,768,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,512,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,128,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,6144,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,512,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,64,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,768,32,0.0021137777302000257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,256,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,4096,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,65536,0.02502222193611993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,128,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,51200,0.022671999202834234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,64,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,65536,0.056877334912618004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,16384,0.010739555789364709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,65536,0.05888977977964613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,12288,0.009036444127559662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,51200,0.04683644572893778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,16384,0.016647110382715862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,10240,0.011162666810883416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,16384,0.01789422167672051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,12288,0.014069333672523499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,128,0.0023555554863479403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,12288,0.015003555350833468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,8192,0.0099973330895106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,7168,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,10240,0.01273333364062839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,10240,0.01348622226052814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,8192,0.011077333655622272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,6144,0.008231110870838165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,7168,0.010456889039940305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,5120,0.0074311114019817775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,8192,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,7168,0.010695999695195092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,6144,0.009914666414260864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,4096,0.0070168889231152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,6144,0.010018666585286459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,5120,0.009332444104883406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,3584,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,128,0.005868444426192178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,4096,0.008640000389681922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,5120,0.009034666750166152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,3072,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,2560,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,4096,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,32,0.005133333305517833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,3072,0.007687999970383114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,3584,0.007699555820888943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,2048,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,51200,0.044857776827282377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,3072,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,1536,0.004208000169859992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,2560,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,2048,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,1024,0.003208000005947219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,2560,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,1536,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,768,0.0032151111712058387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,2048,0.006660444455014334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,512,0.0025617778301239014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,1024,0.006329777754015393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,1536,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,256,0.002505777817633417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,1024,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,128,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,512,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,256,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,64,0.002116444417172008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,128,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,512,32,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,128,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,65536,0.018594667315483093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,64,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,32,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,3584,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,51200,0.01571733256181081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,65536,0.050541334682040744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,65536,0.0574551092253791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,16384,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,51200,0.04005422194798788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,51200,0.045963555574417114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,12288,0.009293333523803288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,16384,0.016511999898486666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,10240,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,16384,0.01756266587310367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,12288,0.014093332820468478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,8192,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,10240,0.012610666453838348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,12288,0.014498665928840637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,768,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,7168,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,512,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,10240,0.013154666456911298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,6144,0.006895111252864202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,256,0.004966222163703707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,8192,0.011575111084514193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,5120,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,7168,0.01037333326207267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,4096,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,7168,0.010682666467295753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,6144,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,3584,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,6144,0.010049777726332346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,3072,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,4096,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,5120,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,2560,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,4096,0.008006222546100616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,3584,0.007770666645632849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,3072,0.007703111403518253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,2048,0.004231111043029361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,2560,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,3072,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,1536,0.003855111284388436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,2560,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,2048,0.007185777856243982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,1024,0.0031759999692440033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,2048,0.006515555497672822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,1536,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,768,0.002814222127199173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,1024,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,1536,0.006167999986145232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,512,0.002521777732504739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,768,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,1024,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,256,0.002199999988079071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,512,0.006091555373536215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,128,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,256,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,512,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,256,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,256,32,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,5120,0.009325332939624786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,128,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,65536,0.014667555689811707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,3584,0.00794666674402025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,64,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,51200,0.012717333104875354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,32,0.00508355556262864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,65536,0.04919911093182034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,16384,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,65536,0.0554986662334866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,51200,0.0395342210928599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,12288,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,51200,0.044509334696663745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,16384,0.01753155555990007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,10240,0.006656888872385025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,12288,0.013819555441538492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,12288,0.014510222607188754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,8192,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,768,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,10240,0.012753778033786349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,8192,0.011075555450386472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,10240,0.01312355531586541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,7168,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,6144,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,8192,0.011077333655622272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,128,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,8192,0.011359111302428775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,5120,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,7168,0.010454222559928894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,6144,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,4096,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,6144,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,5120,0.009007111191749573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,3584,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,4096,0.008347555167145198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,5120,0.008657777474986183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,3072,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,4096,0.007939555578761632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,3584,0.007985778152942657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,2560,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,3584,0.00755555596616533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,16384,0.016695999436908297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,3072,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,2048,0.004200889004601372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,1536,0.0035662220584021676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,2560,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,3072,0.007110222346252865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,1024,0.00314666661951277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,2048,0.006989333364698622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,2560,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,768,0.0027724444452259275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,1536,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,2048,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,1536,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,512,0.002510222295920054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,768,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,1024,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,256,0.0021688888470331826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,512,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,768,0.005508444375462002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,128,0.0018915554715527429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,7168,0.010666666759385003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,64,0.0017964444640609955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,128,32,0.001800000046690305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,256,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,128,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,65536,0.017512889371977914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,128,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,51200,0.015020444989204407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,16384,0.008307555483447181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,64,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,12288,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,32,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,65536,0.05483377642101712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,51200,0.04404000110096402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,16384,0.01741333305835724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,8192,0.006575999988449945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,7168,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,12288,0.014491554763582019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,6144,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,10240,0.01311288856797748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,5120,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,4096,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,8192,0.011396444506115384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,7168,0.01038755559258991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,3584,0.005879999862776862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,6144,0.009602666729026372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,3072,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,2560,0.004856889032655292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,5120,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,2048,0.004209777961174647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,512,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,3584,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,1536,0.003524444583389494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,1024,0.0028862222615215513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,3072,0.007045333584149678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,768,0.002875555513633622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,2560,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,2048,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,512,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,256,0.0021271111650599376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,1536,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,128,0.0018728888697094387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,768,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,10240,0.00647199981742435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,64,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,512,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,64,32,0.0018177777528762817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,256,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,65536,0.01789955629242791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,128,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,51200,0.01514488955338796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,64,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,32,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,16384,0.008430221842394935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,12288,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,65536,0.05464533302519056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,10240,0.007071110937330458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,51200,0.04383644461631775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,8192,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,4096,0.007897777689827813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,7168,0.006309333360857434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,12288,0.01442311041884952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,6144,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,10240,0.012862222062216865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,5120,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,8192,0.011409777734014722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,7168,0.010417777631017896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,6144,0.009670221971140968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,1024,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,3584,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,5120,0.00870577742656072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,3072,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,2560,0.004600000050332812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,4096,0.007715555528799693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,2048,0.0041848888827694785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,3584,0.007491555478837755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,3072,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,1536,0.003519999898142285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,1024,0.0028773333049482773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,1024,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,2048,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,768,0.0028391112056043413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,512,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,1536,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,1024,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,256,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,128,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,16384,0.01756800048881107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,64,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,768,0.0052915554907586836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,512,0.005283555636803309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,256,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,128,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,64,0.005141333159473208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,32,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,2560,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,16384,0.4040044413672553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,16384,0.3397848871019151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,51200,1.2755164040459526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,51200,1.094056871202257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,225,32,4096,0.006567111031876669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,16384,0.14174311690860325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,12288,0.3085262245602078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,51200,0.4232657750447591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,10240,0.26908087730407715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,12288,0.10741688807805379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,12288,0.26000089115566677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,10240,0.21278844939337838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,8192,0.21959022680918375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,10240,0.09060088793436687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,8192,0.17439466052585176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,7168,0.1984435584810045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,8192,0.07397422525617811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,16384,0.3104720115661621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,12288,0.23561867078145346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,7168,0.15145599842071533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,7168,0.06563999917772081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,6144,0.16269066598680285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,10240,0.19863288932376436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,5120,0.13804266187879774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,6144,0.13179732693566218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,6144,0.05735288725958931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,8192,0.16457688808441162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,5120,0.1118897729449802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,4096,0.11761155393388535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,7168,0.142976893319024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,5120,0.04842933350139194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,3584,0.09956266482671101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,4096,0.09389422337214152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,6144,0.12201955583360459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,3072,0.08668444554011027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,3584,0.036871112055248685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,4096,0.08535110950469971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,5120,0.10362311204274495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,2560,0.07224622037675646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,3584,0.07642933395173815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,3072,0.03278222348954942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,3072,0.0686746637026469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,2048,0.05858400132921007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,2560,0.059828446971045605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,2560,0.028957333829667833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,3072,0.06760888629489474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,1536,0.045956446064843066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,2048,0.05070222086376614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,2048,0.025238222546047632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,2560,0.05801066425111559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,1024,0.03297244509061178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,4096,0.04125688804520501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,1536,0.01958755486541324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,1536,0.039701332648595176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,2048,0.05006666647063362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,768,0.025660443637106154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,3584,0.07980977826648288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,1024,0.015561777684423657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,1536,0.04276000128851997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,1024,0.031545778115590416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,512,0.01851733360025618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,768,0.014105778601434497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,768,0.02532355487346649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,1024,0.03440977798567878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,256,0.010612444745169746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,512,0.013251555462678274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,768,0.03011466728316413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,512,0.021480000681347314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,128,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,256,0.01899466746383243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,256,0.012006222373909421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,64,0.00794666674402025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,512,0.025403555896547105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,32,0.008088000118732452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,128,0.01833866702185737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,128,0.011544889046086205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,256,0.02107644412252638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,51200,0.9978977839152018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,64,0.017975111802419026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,128,0.016978666186332703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,32,0.01790488925245073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,51200,0.868577798207601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,65536,1.1258808771769206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,65536,1.0803751415676541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,51200,0.8529218037923177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,16384,0.28732177946302623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,65536,0.42345065540737575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,16384,0.11219288243187799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,16384,0.280074675877889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,12288,0.21595288647545707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,51200,0.33531912167867023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,10240,0.17269955741034615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,12288,0.22960623105367026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,12288,0.08527644475301106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,10240,0.17390133274926078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,8192,0.16095022360483804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,10240,0.07176178031497531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,16384,0.25342488288879395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,8192,0.14194666014777288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,8192,0.05894222524431017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,7168,0.12659022543165419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,12288,0.19115377797020805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,6144,0.11091822385787964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,7168,0.11849244435628255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,7168,0.05253155695067512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,10240,0.16525333457522923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,8192,0.132076448864407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,5120,0.09645066658655803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,6144,0.045531554354561694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,6144,0.10440978076722886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,5120,0.08913599782519871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,7168,0.11619999673631455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,6144,0.10054133335749309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,5120,0.03887288769086202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,5120,0.08683288759655422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,3584,0.07245333327187432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,4096,0.033329778247409396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,4096,0.0702435572942098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,3584,0.0644542243745592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,3072,0.0626479983329773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,3584,0.02996888756752014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,3584,0.06352444489796956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,2560,0.05167466733190748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,3072,0.026907554931110803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,3072,0.05631555451287163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,51200,0.7688088946872287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,2048,0.04159733321931627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,4096,0.07728177971310086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,2560,0.023968888653649226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,2560,0.04774577750100029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,1536,0.033129778173234724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,2048,0.02010933392577701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,2048,0.045662220981385976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,4096,0.0729315545823839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,1024,0.02429155508677165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,2560,0.05008444521162245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,1536,0.0327413347032335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,1536,0.018420444594489205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,768,0.01863911085658603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,1024,0.013895110951529609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,1024,0.025729777084456548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,1536,0.036899556716283165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,512,0.012527999778588613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,768,0.012778666284349231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,1024,0.03045066528850132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,65536,1.021916495429145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,768,0.02137600051032172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,256,0.011343110766675739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,512,0.011725333001878528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,512,0.017858665850427415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,3072,0.05639644463857015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,128,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,256,0.016083555089102853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,256,0.010445333189434476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,256,0.019519999623298645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,128,0.010589333044158088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,128,0.015062222878138224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,32,0.006535111202134027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,64,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,2048,0.04429866539107429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,128,0.015000000596046448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,64,0.015062222878138224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,32,0.015119999647140503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,65536,0.38722756173875594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,51200,0.3042079872555203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,65536,0.35724443859524196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,768,0.026408889227443274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,65536,0.1455119980706109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,16384,0.10675111081865098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,51200,0.28323109944661456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,512,0.023401778605249193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,16384,0.04071377714474996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,12288,0.07624177800284491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,51200,0.11491910616556804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,16384,0.08896266751819187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,10240,0.06754933463202582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,12288,0.06990755266613431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,12288,0.03234577841228909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,12288,0.07059911224577162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,10240,0.05898577637142605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,10240,0.02811644474665324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,8192,0.05237244566281637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,10240,0.060848891735076904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,65536,0.33368621932135684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,8192,0.023690667417314317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,8192,0.049230221245023936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,7168,0.04717777834998237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,51200,0.26922843191358775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,6144,0.042282667424943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,16384,0.0922408898671468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,7168,0.02149155570401086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,7168,0.045137776268853076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,5120,0.03638133406639099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,8192,0.05181688732571072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,4096,0.030116445488399927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,6144,0.03820711043145921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,7168,0.047239998976389565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,5120,0.03298577666282654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,3584,0.026499556170569524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,4096,0.028070221344629925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,4096,0.01514400045077006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,5120,0.03784888982772827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,3072,0.023008000519540574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,4096,0.032411555449167885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,3584,0.013790222505728403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,3584,0.02494666642612881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,2560,0.01937955617904663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,3072,0.012810666528013019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,3584,0.03021777669588725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,3072,0.02204888893498315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,2048,0.015824889143308003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,2560,0.019338667392730713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,2560,0.012106666962305704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,3072,0.027911110056771174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,1536,0.012648888760142855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,6144,0.018776888648668926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,2560,0.02436622149414486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,2048,0.01696266730626424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,1024,0.009935110807418823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,5120,0.016923555069499545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,1536,0.014099554883109199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,1536,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,6144,0.042230221960279674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,768,0.008319999608728621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,1024,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,1024,0.011791110866599612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,1536,0.01959911154376136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,512,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,768,0.010268444816271463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,1024,0.016551999582184684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,768,0.015204444527626038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,512,0.009248889154858058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,512,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,512,0.01313244468635983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,256,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,128,0.0038951109680864546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,256,0.008614222208658854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,256,0.011791999969217511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,128,0.006711999989218182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,128,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,64,0.003536000019974179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,32,0.003485333174467087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,128,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,2048,0.021969777014520433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,64,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,32,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,65536,0.2943528758154975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,768,0.00741244438621733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,51200,0.23545598983764648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,65536,0.29269777403937447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,256,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,65536,0.1283759938346015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,16384,0.07719288931952582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,51200,0.22679021623399523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,51200,0.1013902227083842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,16384,0.03825689024395413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,16384,0.07698577642440796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,2048,0.011044444309340583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,16384,0.08679644266764323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,12288,0.030119998587502375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,10240,0.051177776522106595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,12288,0.05865244732962715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,12288,0.06867200136184692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,10240,0.02565244502491421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,65536,0.31579287846883136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,8192,0.04394488864474826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,51200,0.2466648949517144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,7168,0.03632444474432204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,10240,0.060030221939086914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,8192,0.021800888909233943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,8192,0.04222222169240316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,6144,0.03161599900987413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,8192,0.04975377851062351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,7168,0.036941332949532404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,12288,0.06312533219655354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,5120,0.027719111906157598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,7168,0.045824001232783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,6144,0.03260355525546604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,6144,0.018079999420377944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,4096,0.023853333459960088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,5120,0.016142222616407607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,6144,0.04055644406212701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,5120,0.028044445647133723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,3584,0.020543111695183646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,4096,0.014582221706708273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,10240,0.05049066742261251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,5120,0.036299556493759155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,4096,0.03147466646300422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,3584,0.02136444383197361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,3072,0.012488000094890594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,3584,0.028554668029149372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,2560,0.015125332607163323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,7168,0.019825778073734708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,3072,0.026312889324294195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,2048,0.01295644458797243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,2560,0.017102221647898357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,2560,0.02386755579047733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,2048,0.014889778362380134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,1536,0.010667555862002902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,2048,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,2048,0.021349334054523047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,4096,0.023552000522613525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,1024,0.008614222208658854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,1536,0.012719111310111152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,1536,0.00851644410027398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,3072,0.017813333206706576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,3584,0.013409777647919126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,768,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,1024,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,3072,0.019084443648656208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,512,0.0058702222175068325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,768,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,768,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,1024,0.01517599986659156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,256,0.0042053332759274375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,2560,0.011712888876597086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,512,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,512,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,128,0.003530666646030214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,256,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,256,0.0074577778577804565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,128,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,64,0.003489777859714296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,128,0.009662222531106737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,32,0.0035137778355015647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,64,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,1536,0.019183999962276883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,1024,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,32,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,65536,0.23393334282769096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,768,0.014007111390431723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,65536,0.24747112062242296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,65536,0.1245759990480211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,512,0.012776888906955719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,16384,0.0704995526207818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,51200,0.1934373378753662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,51200,0.099053336514367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,16384,0.06581333610746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,16384,0.03712266683578491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,12288,0.04998044504059685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,16384,0.08636266655392116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,12288,0.05089244577619764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,65536,0.3121946652730306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,10240,0.04402933186954922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,51200,0.18907645013597277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,12288,0.028632889191309612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,51200,0.24424354235331217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,8192,0.03399466805987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,12288,0.06810222069422404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,10240,0.024600888291994732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,10240,0.04323644439379374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,7168,0.03094666533999973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,8192,0.03639199998643663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,8192,0.02125866711139679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,128,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,10240,0.05748889181349012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,6144,0.025972444150182936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,7168,0.01919555498494042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,8192,0.04987288845909966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,7168,0.03222577770551046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,5120,0.022368000613318548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,6144,0.017553778158293832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,6144,0.028377778000301782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,7168,0.04518844352828132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,5120,0.015750222735934787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,5120,0.024147555232048035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,6144,0.040525333748923406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,3584,0.016131555040677387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,4096,0.014112888111008538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,4096,0.020635555187861126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,5120,0.03509689039654202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,3072,0.014561777313550314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,3584,0.013100444442696042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,3584,0.01865688959757487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,4096,0.03046311272515191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,2560,0.012888888518015543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,3072,0.01187999960449007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,3584,0.027503111296229895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,256,0.011720000041855706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,2048,0.011212444139851464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,2560,0.015280889140235053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,2560,0.010703110860453712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,3072,0.02567555507024129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,1536,0.008999110923873054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,2048,0.013399110900031196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,2048,0.009124444590674506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,4096,0.019655999210145738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,1024,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,2560,0.022846221923828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,1536,0.011671110987663269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,1536,0.008064889245563084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,768,0.00619555554456181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,1024,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,1024,0.00907733374171787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,512,0.004701333327425851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,768,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,1536,0.01866666641500261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,1024,0.015102222561836243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,768,0.008280889027648503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,256,0.003866666720973121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,512,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,512,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,768,0.013792888985739814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,128,0.0032711111836963226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,3072,0.017091555727852713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,256,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,64,0.0031964443624019623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,256,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,512,0.012691555751694573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,32,0.003540444291300244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,128,0.006649777707126405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,128,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,256,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,65536,0.19618222448560926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,128,0.009317333499590555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,32,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,65536,0.07733422517776489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,51200,0.15054666996002197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,65536,0.20306666692097983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,2048,0.020982222424613103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,16384,0.050832000043657094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,51200,0.15950489044189453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,12288,0.0424560010433197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,51200,0.061956445376078285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,16384,0.023927999867333308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,16384,0.05443200137880114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,10240,0.033498668008380465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,12288,0.04247022337383694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,12288,0.0192586663696501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,65536,0.17830310927497017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,51200,0.14000888665517172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,8192,0.029184887806574505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,10240,0.036293334431118436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,10240,0.016955556141005624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,12288,0.04284800092379252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,7168,0.025054223007626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,64,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,8192,0.015040889382362366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,8192,0.03038311004638672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,10240,0.036848889456854925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,6144,0.021645334031846788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,7168,0.026774222652117412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,7168,0.013791999883121915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,8192,0.03250577714708116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,5120,0.018826666805479262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,6144,0.012865777644846173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,4096,0.015729778342776828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,6144,0.023659555448426142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,5120,0.011755555868148804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,5120,0.020610666937298246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,6144,0.026774222652117412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,3584,0.013714666995737286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,4096,0.010781333678298526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,4096,0.01791822248035007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,16384,0.05249333381652832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,4096,0.021512000097168818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,3584,0.016341333587964375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,3584,0.009834667046864828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,3072,0.012107555237081317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,2560,0.010455999937322406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,3072,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,3584,0.019932443896929424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,3072,0.015022221538755627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,2560,0.013549333645237817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,2560,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,3072,0.019139554765489366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,2048,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,2560,0.017614222235149808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,2048,0.011867555479208628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,1536,0.0075937774446275495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,7168,0.029623111089070637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,1024,0.006214222146405115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,2048,0.01622577839427524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,1536,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,1536,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,768,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,5120,0.023736889163653057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,1024,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,1536,0.014722666806644864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,1024,0.00832088871134652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,512,0.004201777693298128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,768,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,1024,0.012423111332787408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,256,0.0031946665710873077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,512,0.006218666831652324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,512,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,768,0.011672000090281168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,128,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,2048,0.009299555586444007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,256,0.005967111223273807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,256,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,64,0.0026968889352348116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,32,0.0028364445186323593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,128,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,128,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,256,0.010041777458455827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,64,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,65536,0.18131466706593832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,32,0.006559111177921295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,51200,0.14165600140889487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,65536,0.19730044735802543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,65536,0.07283111413319905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,16384,0.052482667896482676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,51200,0.15443110466003418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,768,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,51200,0.05919644567701551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,12288,0.04013244311014811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,16384,0.05337599913279215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,16384,0.02238311039076911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,65536,0.16779288980695936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,512,0.011019555230935415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,10240,0.033732444047927856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,16384,0.05061155557632446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,12288,0.04125510983996921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,51200,0.13198578357696533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,8192,0.027836445305082533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,10240,0.016523554921150208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,10240,0.0351528889603085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,12288,0.04214844438764784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,128,0.008408889174461365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,8192,0.014589332871966891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,8192,0.0292871097723643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,10240,0.036907556984159685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,6144,0.020869332883093093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,8192,0.03158844510714213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,7168,0.026013332936498854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,5120,0.017846221725145977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,6144,0.0233608881632487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,6144,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,7168,0.02870222263866001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,4096,0.015135111080275642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,5120,0.011429333024554782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,5120,0.020253333780500624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,6144,0.02659822172588772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,3584,0.013793778088357715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,4096,0.01001422190003925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,4096,0.017474666237831116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,5120,0.023447111248970032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,3072,0.011861333416567909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,3584,0.01592266725169288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,3584,0.009308444129096137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,4096,0.021185777253574792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,2560,0.010418666733635796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,7168,0.024241778585645888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,3072,0.01465777721669939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,3072,0.00850133349498113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,3584,0.019885333047972787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,2048,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,7168,0.013440888788965015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,3072,0.018464000688658822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,1536,0.007870222131411234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,2048,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,2048,0.0073831114504072405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,2560,0.017344888713624742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,2048,0.015814221567577787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,1536,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,1024,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,1536,0.009479111267460717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,1024,0.00666222224632899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,12288,0.018579555882347953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,1536,0.013463999662134381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,768,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,1024,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,512,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,768,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,1024,0.011863999896579318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,768,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,256,0.003839111162556542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,512,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,512,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,768,0.011121778024567498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,128,0.00313688897424274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,2560,0.007822222179836696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,256,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,512,0.010671999719407825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,256,0.00645955569214291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,64,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,32,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,128,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,128,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,256,0.010690666735172272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,64,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,65536,0.16503466500176325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,128,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,51200,0.12991110483805338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,65536,0.15981688764360216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,65536,0.07076978021197848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,16384,0.04807733496030172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,51200,0.12461600038740371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,51200,0.0573128859202067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,16384,0.04407377706633674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,65536,0.16242310735914442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,12288,0.03712266683578491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,16384,0.021573333276642695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,51200,0.1281928883658515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,16384,0.05130666825506422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,12288,0.017975999249352347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,10240,0.030395554171668157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,12288,0.034068445364634194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,8192,0.025643555654419795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,10240,0.015899555550681222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,32,0.006367110957702001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,12288,0.04132089018821716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,7168,0.022315555148654517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,8192,0.024336889386177063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,8192,0.014101333088344999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,10240,0.03626044591267904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,6144,0.020302222834693063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,7168,0.022043555974960327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,7168,0.013068444199032254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,5120,0.016362667083740234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,6144,0.012190221912331052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,7168,0.02797777785195245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,6144,0.025757332642873127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,2560,0.01315199997689989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,5120,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,5120,0.023279110590616863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,4096,0.014919110470347934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,3584,0.012472000386979846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,4096,0.010000000397364298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,4096,0.020675554871559143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,10240,0.028917332490285236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,3072,0.010654222634103564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,3584,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,3584,0.013830222189426422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,3072,0.012934221989578672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,3072,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,3584,0.019581332802772522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,6144,0.0196435550848643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,3072,0.018232888645595975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,5120,0.017224889662530687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,2048,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,4096,0.013909333282046847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,2560,0.0169813334941864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,2048,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,1536,0.0070186663005087115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,2048,0.009931555224789513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,2048,0.014939554863505893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,1536,0.0069164443347189165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,1024,0.005369777894682354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,1536,0.008420444197124904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,1536,0.01354311075475481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,768,0.0042151109211974675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,1024,0.00738666703303655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,2560,0.009654222263230218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,1024,0.011759111450778114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,512,0.0035182221068276297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,8192,0.03076710965898302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,768,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,768,0.006312888943486744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,2560,0.011414222419261932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,2560,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,256,0.002862222285734283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,512,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,768,0.011384000380833944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,128,0.002597333242495855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,256,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,512,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,256,0.006096000058783426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,64,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,128,0.005971555494599872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,32,0.0025457777082920074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,128,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,64,0.006127999888526068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,128,0.008012444608741337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,1024,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,32,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,51200,0.11781956089867486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,65536,0.06789511442184448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,65536,0.1546622249815199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,51200,0.11992177698347305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,512,0.006311999840868845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,51200,0.054808888170454234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,16384,0.041505777173572116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,65536,0.16210133499569362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,12288,0.03333866596221924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,16384,0.020972443951500788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,51200,0.12882667117648655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,16384,0.041866666740841336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,256,0.01032266683048672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,10240,0.027504000398847792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,12288,0.032701333363850914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,12288,0.017401778035693698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,16384,0.05056088831689623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,65536,0.14649066660139295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,8192,0.024363555841975745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,10240,0.02812533246146308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,7168,0.01942755613062117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,8192,0.013809777796268463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,12288,0.04008355405595567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,8192,0.02385777731736501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,10240,0.03529511226548089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,7168,0.012792000340090858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,7168,0.021475555168257818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,8192,0.030356443590588037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,5120,0.015797333584891427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,7168,0.0280728903081682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,6144,0.019208000765906442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,5120,0.01035199976629681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,6144,0.025639111797014873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,5120,0.01706755492422316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,4096,0.013454222016864352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,3584,0.011863999896579318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,4096,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,5120,0.022775999373859827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,4096,0.014819555812411837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,3072,0.010294222169452244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,3584,0.013753777576817406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,3584,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,10240,0.01552088889810774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,4096,0.020581333173645865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,2560,0.009751111268997192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,3072,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,3072,0.012458666331238218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,6144,0.017465778523021273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,3584,0.019335110982259113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,2048,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,2560,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,2560,0.011239111423492432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,1536,0.007254222200976477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,3072,0.018191110756662156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,2048,0.007287110719415877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,2560,0.01686844395266639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,2048,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,6144,0.011426666544543372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,1536,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,768,0.0042035554846127825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,1024,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,1024,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,1536,0.01329422245422999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,512,0.0035582222044467926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,768,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,1024,0.01202666676706738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,768,0.006302222195598815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,256,0.003157333367400699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,512,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,512,0.006054222169849608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,768,0.011398221883508893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,128,0.0028124445428450904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,256,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,256,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,512,0.01034222212102678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,64,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,128,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,128,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,32,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,1024,0.005417777846256892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,65536,0.10868799686431885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,64,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,128,0.008049777812427944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,32,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,51200,0.09375555647744073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,2048,0.014704000618722705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,65536,0.06539022260242038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,65536,0.11456799507141113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,16384,0.03549422158135308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,51200,0.09075733025868733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,51200,0.052480889691246875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,65536,0.15257155895233154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,16384,0.031814220878813006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,16384,0.020346666375796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,51200,0.12184622552659775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,16384,0.04784799946679009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,12288,0.01679111189312405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,12288,0.02513244416978624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,256,0.009656888743241629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,10240,0.024527110987239416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,8192,0.020593777298927307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,10240,0.015127110812399121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,12288,0.03830488853984409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,1536,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,10240,0.02223644488387638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,7168,0.017779555585649278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,8192,0.018975110517607797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,8192,0.013354666531085968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,6144,0.015559110376569959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,7168,0.0120666664507654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,7168,0.01718755563100179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,8192,0.02892355455292596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,5120,0.013376000026861826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,6144,0.015471110741297403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,6144,0.011071999867757162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,12288,0.029383109675513372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,4096,0.012183999849690331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,5120,0.010016889207892949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,5120,0.013790222505728403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,6144,0.024404444628291663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,3584,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,4096,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,5120,0.021987555755509272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,4096,0.012077333198653327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,3072,0.00814844419558843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,3584,0.011552000211344825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,3584,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,4096,0.01958044370015462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,2560,0.007159999675220913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,3072,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,3584,0.018434666925006442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,10240,0.03335111008750068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,2048,0.006227555374304454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,2560,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,3072,0.016864000095261466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,1536,0.005364444520738389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,2560,0.009014222357008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,7168,0.026384888423813715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,2560,0.01551999979548984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,1024,0.004253333227501975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,2048,0.013808888693650564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,1536,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,1536,0.007339555356237624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,768,0.0035573335157500375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,1024,0.006435555716355641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,1536,0.012704888979593912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,1024,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,512,0.003156444471743372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,768,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,768,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,1024,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,256,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,512,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,512,0.005967111223273807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,3072,0.007962666451931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,128,0.002483555633160803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,256,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,256,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,512,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,64,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,128,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,2048,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,128,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,256,0.009645333720578088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,64,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,65536,0.09855733315149943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,128,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,32,0.0058017778727743365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,51200,0.0852844450208876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,65536,0.11513777573903401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,65536,0.06436800294452243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,16384,0.03242311212751601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,51200,0.08996622429953681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,51200,0.05131733417510986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,65536,0.15149511231316462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,12288,0.027930667002995808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,768,0.011141333315107556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,16384,0.020039111375808716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,51200,0.12071200211842854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,10240,0.02355466617478265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,12288,0.01664177742269304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,12288,0.025220443805058796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,16384,0.046889778640535146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,2048,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,8192,0.019254222512245178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,10240,0.021864000293943617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,10240,0.014893333117167154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,12288,0.03791022300720215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,7168,0.016964443855815463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,8192,0.01309866706530253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,8192,0.018926221463415358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,10240,0.03267555435498556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,6144,0.014859555496109856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,7168,0.01184800018866857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,8192,0.02830044428507487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,6144,0.010843555960390302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,5120,0.013038222160604266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,7168,0.026143999563323125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,16384,0.03193333413865831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,6144,0.024195555183622573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,4096,0.010610666539933948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,5120,0.013656888571050433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,32,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,5120,0.021753778060277302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,3584,0.009824000298976898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,4096,0.012082666986518435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,4096,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,4096,0.019526221685939364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,3584,0.008340444829728868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,3584,0.011134222149848938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,7168,0.017229333519935608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,3584,0.018232888645595975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,2560,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,3072,0.010001777774757808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,3072,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,3072,0.016181333197487723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,2560,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,2048,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,2560,0.008972444468074376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,5120,0.009373333719041612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,1536,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,2048,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,2560,0.014986667368147107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,2048,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,1024,0.004587555511130227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,2048,0.013808888693650564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,1536,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,1536,0.007276444799370236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,768,0.0041680000722408295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,1024,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,1024,0.0064106666379504735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,512,0.003546666767862108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,768,0.006228444476922353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,768,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,1024,0.011794666449228922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,256,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,512,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,768,0.01072177787621816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,6144,0.01550222271018558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,128,0.0024560000747442245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,256,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,256,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,512,0.010072888599501716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,64,0.0022168887986077205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,128,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,128,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,32,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,256,0.009519110951158736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,65536,0.08335377772649129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,128,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,64,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,32,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,51200,0.06748711400561862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,65536,0.10916799969143337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,3072,0.008903111020723978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,16384,0.02860355708334181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,51200,0.050182223320007324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,51200,0.08594222201241387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,65536,0.15137333340115017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,512,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,12288,0.023943111300468445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,16384,0.019817777805858187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,16384,0.03053955568207635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,51200,0.12043110529581706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,10240,0.020964443683624268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,12288,0.016529776983790927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,12288,0.0247715562582016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,16384,0.046219554212358266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,8192,0.016583111551072862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,10240,0.014734221829308404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,10240,0.02154133386082119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,7168,0.015180443723996481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,12288,0.03747111227777269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,8192,0.01236533373594284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,8192,0.018418666389253404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,10240,0.032636443773905434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,6144,0.014019555515713163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,7168,0.010955555571450127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,7168,0.016757332616382174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,8192,0.028258668051825628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,1536,0.012706666356987424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,6144,0.01514755520555708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,6144,0.010055999788973067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,4096,0.009815110928482479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,7168,0.026169778572188482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,5120,0.01329422245422999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,6144,0.02379111117786831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,5120,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,3584,0.007991111112965478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,4096,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,4096,0.011400000088744693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,3584,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,3072,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,4096,0.018586667047606576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,3584,0.01723377737734053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,3072,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,2560,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,3072,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,3072,0.016490666402710807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,5120,0.011570666399266986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,2048,0.0058408888677755995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,2560,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,2560,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,2560,0.014532445205582513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,2048,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,1536,0.004572444491916233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,5120,0.021320000290870667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,1536,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,1536,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,1024,0.003882666842805015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,3584,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,1536,0.012431110772821637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,768,0.003378666730390655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,1024,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,1024,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,1024,0.010848888920413123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,768,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,768,0.006292444550328785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,65536,0.06323999828762479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,768,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,512,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,256,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,512,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,2048,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,128,0.0022657778528001574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,512,0.010016000105275048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,256,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,64,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,256,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,32,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,128,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,2048,0.013803555733627744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,128,0.0053297777970631914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,256,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,65536,0.0771991080707974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,64,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,51200,0.05900000201331245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,128,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,32,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,16384,0.02416444487041897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,65536,0.06232444445292155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,65536,0.10589510864681667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,51200,0.04898311032189263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,51200,0.08366222514046563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,12288,0.02023555503951179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,512,0.002883555574549569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,16384,0.0196088883611891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,16384,0.029826667573716905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,65536,0.15100267198350695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,10240,0.017698667115635343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,51200,0.12035555309719509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,8192,0.014223999447292753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,10240,0.014115555418862237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,12288,0.02404800057411194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,10240,0.02088533341884613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,7168,0.01274400038851632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,12288,0.03711822297837999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,8192,0.011697777443461947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,10240,0.03231733375125461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,6144,0.011209777659840055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,7168,0.010910222099887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,7168,0.016398222910033334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,8192,0.028301331732008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,5120,0.009830222361617619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,6144,0.010019555687904358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,6144,0.014747555057207743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,5120,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,5120,0.012457778056462606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,6144,0.02365600069363912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,4096,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,4096,0.008612444831265343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,5120,0.021507556239763897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,3584,0.008721777962313758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,4096,0.01074488874938753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,16384,0.04576977756288317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,3072,0.00815822184085846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,4096,0.018716444571812946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,3584,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,3584,0.010025777750545079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,2560,0.006986666884687211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,8192,0.018056000272432964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,3072,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,3584,0.017217778497272067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,3072,0.009311999711725447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,2048,0.006224888894293044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,2560,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,2560,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,3072,0.015909334023793537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,1536,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,7168,0.02572088936964671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,2048,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,1536,0.007198221981525421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,1536,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,1024,0.00424799985355801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,1536,0.012148444851239523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,1024,0.0063386667105886675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,1024,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,1024,0.010753778119881948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,768,0.005974222388532426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,768,0.006167999986145232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,512,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,768,0.010477333433098264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,2560,0.014576888746685453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,512,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,512,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,256,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,2048,0.013790222505728403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,128,0.0022435554613669715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,12288,0.01628444426589542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,256,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,256,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,768,0.003551111039188173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,64,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,128,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,32,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,128,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,65536,0.06155733267466227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,128,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,64,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,32,0.005633777628342311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,51200,0.05043288734224108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,65536,0.06645333104663424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,65536,0.06041510899861654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,16384,0.0210817770825492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,2048,0.007856888903511895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,51200,0.052601777844958834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,51200,0.047446221113204956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,65536,0.1505662202835083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,12288,0.0177484436167611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,16384,0.01923644377125634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,16384,0.02036888897418976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,51200,0.1200444433424208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,10240,0.017239999439981248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,512,0.009777777724795872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,12288,0.015457777513398064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,8192,0.013471110827393003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,16384,0.045681777927610606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,10240,0.01386577801571952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,12288,0.03663466705216302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,256,0.009071110851234859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,7168,0.011917333636018964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,8192,0.011407111254003314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,8192,0.013061333033773633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,6144,0.010683555569913654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,7168,0.010860444770918952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,7168,0.012050666742854647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,8192,0.027504889501465693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,5120,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,6144,0.010683555569913654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,6144,0.009993777506881291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,5120,0.009310222334331935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,4096,0.008285333712895712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,5120,0.0092275556590822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,12288,0.01680177781316969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,5120,0.020070221689012315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,3584,0.007432888779375289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,4096,0.00833511104186376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,4096,0.008611555728647444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,10240,0.014849777022997538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,4096,0.018238221605618794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,3072,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,3584,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,10240,0.03208444515864054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,2560,0.00600000015563435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,3072,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,3584,0.016528000434239704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,3072,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,2048,0.005095110999213325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,2560,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,3072,0.015878222054905362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,2560,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,7168,0.025555556019147236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,1536,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,2048,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,2048,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,1024,0.0035253332720862497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,2560,0.014364444547229342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,6144,0.023281777898470562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,1536,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,1536,0.006315555423498154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,768,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,2048,0.01311644415060679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,1024,0.006313777632183499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,1536,0.012131555212868584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,1024,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,768,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,768,0.005649777750174205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,1024,0.011061333119869232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,256,0.002492444382773505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,512,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,768,0.010649777948856354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,512,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,128,0.0024862223201327855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,256,0.005873777800136142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,256,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,128,0.0057555557125144536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,256,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,128,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,32,0.0021928888228204516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,64,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,65536,0.0491848886013031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,512,0.002798222212327851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,32,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,65536,0.06404711140526666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,3584,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,65536,0.0588302214940389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,51200,0.03956088754865859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,512,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,65536,0.14975733227199978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,16384,0.017113778326246474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,51200,0.05028355452749464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,128,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,64,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,12288,0.01481066644191742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,16384,0.018544889158672757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,51200,0.11896533436245388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,10240,0.01352533366945055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,12288,0.01516533394654592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,16384,0.044848889112472534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,10240,0.014454222387737699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,10240,0.012901333471139273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,12288,0.03656355539957682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,8192,0.013111111190583972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,10240,0.03161066770553589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,8192,0.011371555427710215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,7168,0.012610666453838348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,8192,0.01202311118443807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,7168,0.010670222342014313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,8192,0.02750933335887061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,51200,0.0468355549706353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,7168,0.011390222443474663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,6144,0.011979555918110741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,5120,0.00923288861910502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,6144,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,7168,0.024540444215138752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,16384,0.019842666056421068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,4096,0.007605333295133378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,5120,0.009317333499590555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,6144,0.022426666484938726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,12288,0.01649777756796943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,3584,0.007128888534175024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,4096,0.008656000097592672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,4096,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,3072,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,3584,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,3584,0.008226667013433244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,4096,0.01789955629242791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,2560,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,3584,0.016547555724779766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,3072,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,2560,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,2560,0.006991110742092133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,2048,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,3072,0.015448000695970325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,2048,0.007175111108356052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,2560,0.014133334159851074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,1536,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,6144,0.010324444207880232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,5120,0.009297777381208207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,1024,0.003847111016511917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,2048,0.01313244468635983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,1536,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,1536,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,768,0.0035386664999855887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,5120,0.019959999455346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,1024,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,1536,0.012099555797047086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,512,0.0031973332580592898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,768,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,1024,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,3072,0.007693332930405934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,256,0.0027946666296985415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,512,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,768,0.010440000229411654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,256,0.0058711109062035876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,128,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,512,0.009872000250551436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,256,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,2048,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,256,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,64,0.0019111111760139465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,32,0.0021520000365045336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,128,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,65536,0.03452355662981669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,128,0.007672888537247975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,64,0.004972444640265571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,1024,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,32,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,51200,0.02892000145382351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,768,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,65536,0.05747555361853706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,65536,0.06146933635075887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,16384,0.013986666997273764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,51200,0.048400001393424145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,512,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,65536,0.14971555603875056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,12288,0.011965333587593503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,16384,0.01717066599263085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,16384,0.019131556153297424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,51200,0.11865066157446967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,10240,0.011415111521879831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,128,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,12288,0.014135110709402295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,16384,0.04527555571662056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,8192,0.010628444453080496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,10240,0.013472889032628802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,12288,0.035990221632851496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,10240,0.01277599980433782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,7168,0.009975999593734741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,8192,0.012099555797047086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,8192,0.011390222443474663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,10240,0.031120889716678198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,6144,0.010016889207892949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,7168,0.010474666953086853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,7168,0.011015999648306104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,8192,0.02644266684850057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,51200,0.046281778150134616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,5120,0.009377777576446533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,6144,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,6144,0.010148444109492833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,7168,0.024060444699393377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,4096,0.007689778175618913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,5120,0.009295110901196798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,6144,0.022295110755496558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,5120,0.009142222503821055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,3584,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,12288,0.01567555632856157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,5120,0.0196106665664249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,3072,0.006790222393141852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,4096,0.01759999990463257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,3584,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,3584,0.007777777810891469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,2560,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,3584,0.0163093325164583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,3072,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,2048,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,2560,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,3072,0.015783111254374187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,2048,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,2560,0.014155555102560254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,1536,0.004301333179076512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,2048,0.00701155596309238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,2048,0.013166222307417126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,1536,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,1024,0.0034968890249729156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,1536,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,4096,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,1536,0.012164444559150271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,1024,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,1024,0.005776888794369168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,768,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,1024,0.010974222587214576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,512,0.002855111120475663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,768,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,3072,0.007743111087216272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,256,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,768,0.010659555594126383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,512,0.005972444597217772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,2560,0.007384888827800751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,512,0.005283555636803309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,128,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,256,0.005034666508436203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,512,0.010363555616802638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,64,0.0021848889688650766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,128,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,32,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,256,0.009741333623727163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,128,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,65536,0.02947910957866245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,64,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,4096,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,32,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,768,0.003157333367400699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,51200,0.024350222614076402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,65536,0.056871996985541455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,65536,0.06092088752322727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,16384,0.014176888598336114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,51200,0.04592000113593208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,51200,0.047556443346871265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,12288,0.012440000143316058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,65536,0.14932089381747776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,16384,0.018942221999168396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,51200,0.11890933248731826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,10240,0.011199110911952125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,16384,0.04519733455446032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,12288,0.015061333775520325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,128,0.005724444571468566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,8192,0.009925333162148794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,10240,0.013468444347381592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,10240,0.030751999881532457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,8192,0.011674666570292579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,7168,0.009278222090668148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,8192,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,8192,0.02666666607062022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,7168,0.01092266705301073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,6144,0.008661333057615492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,7168,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,16384,0.016733333468437195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,7168,0.0240000006225374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,5120,0.007836444510353936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,6144,0.009761778016885122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,12288,0.014127999544143677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,6144,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,4096,0.007189333438873291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,5120,0.008989333278603023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,10240,0.012764444781674279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,6144,0.02204711072974735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,5120,0.009043555292818282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,3584,0.0063662222690052455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,4096,0.008414222134484185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,4096,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,3072,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,5120,0.01957244508796268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,3584,0.007861333588759104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,3584,0.008031110796663497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,2560,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,4096,0.017367111312018502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,3072,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,3584,0.01651822196112739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,3072,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,2048,0.004554666578769684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,2560,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,3072,0.015512888630231222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,1536,0.0041911109454101985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,2048,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,2048,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,2560,0.014126222994592456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,1024,0.00350577798154619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,2048,0.01348444405529234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,1536,0.006664000037643645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,1536,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,768,0.0032071111102898917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,12288,0.03567911187807719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,1024,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,1024,0.005848000033034219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,1536,0.012079111403889127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,512,0.002805333377586471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,768,0.00600622221827507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,768,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,256,0.002499555548032125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,512,0.005306666509972678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,768,0.010677333507272931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,128,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,256,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,512,0.009683555199040307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,256,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,64,0.0021191111041439907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,2560,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,128,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,768,32,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,256,0.009690666364298927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,65536,0.02478933334350586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,64,0.004948444250557157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,128,0.007476444045702617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,32,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,51200,0.018170666363504197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,65536,0.05841244591606987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,65536,0.05709866682688395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,16384,0.010774222513039907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,51200,0.04474577638838026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,65536,0.14780622058444554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,51200,0.046978665722741016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,12288,0.012638222012254926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,1024,0.011412444214026133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,16384,0.016764443781640794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,16384,0.018068444397714403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,512,0.00573333352804184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,51200,0.11790222591824001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,10240,0.011385777758227454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,12288,0.014782221780882942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,12288,0.014112888111008538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,8192,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,16384,0.04480266571044922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,10240,0.0127688886390792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,7168,0.00923733330435223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,10240,0.013166222307417126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,8192,0.011136000355084738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,12288,0.03568177753024631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,128,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,6144,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,8192,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,10240,0.030544890297783747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,5120,0.008014221986134848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,7168,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,7168,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,6144,0.00982488857375251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,8192,0.026064889298544988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,4096,0.00721688899728987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,6144,0.009964444571071202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,7168,0.0236791107389662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,5120,0.009016888837019602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,3584,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,6144,0.021896888812383015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,4096,0.008477333519193862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,4096,0.008012444608741337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,3072,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,5120,0.019596444235907663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,3584,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,3584,0.00777333312564426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,4096,0.01718755563100179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,3072,0.007673777639865875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,3072,0.0073315559162033936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,3584,0.016865778300497267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,2048,0.004855999930037392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,2560,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,2560,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,3072,0.01513777838812934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,1536,0.0042364444169733255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,2048,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,2048,0.006657777975002925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,2560,0.014144889182514615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,1024,0.0032115555885765287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,1536,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,1536,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,2048,0.013092444174819522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,768,0.002993777808215883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,1024,0.0063786668082078295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,1024,0.005691555639108022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,1536,0.011911999848153857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,512,0.002537777854336633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,768,0.005992888990375731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,768,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,1024,0.010775110787815519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,2560,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,512,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,768,0.01091733326514562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,256,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,256,0.005766222046481238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,512,0.010072000324726105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,256,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,128,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,64,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,128,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,32,0.0021439999755885866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,128,0.007301333049933116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,65536,0.018417777286635507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,5120,0.008966222405433655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,64,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,51200,0.014889778362380134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,32,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,65536,0.053245332505967885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,65536,0.05759021970960829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,256,0.002543111021320025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,16384,0.00981244444847107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,512,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,51200,0.03975911272896661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,51200,0.04632711079385546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,12288,0.008406221866607666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,65536,0.14862133396996394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,16384,0.01647911138004727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,10240,0.007899555895063613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,512,128,0.0021271111650599376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,51200,0.1173235575358073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,12288,0.014094221923086377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,12288,0.014599111345079211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,8192,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,16384,0.04336622357368469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,10240,0.012764444781674279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,7168,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,12288,0.034719109535217285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,10240,0.013075555364290873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,8192,0.011086222198274402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,8192,0.01146666705608368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,6144,0.0063893331421746155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,10240,0.029952890343136255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,7168,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,7168,0.010703999963071613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,8192,0.026093333959579468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,6144,0.009707555174827576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,6144,0.010031110710567897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,7168,0.023702222439977858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,4096,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,5120,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,5120,0.008882666627566019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,6144,0.021543999512990315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,3584,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,4096,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,4096,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,5120,0.01961599952644772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,3072,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,3584,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,3584,0.007727999654081132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,16384,0.01777777738041348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,4096,0.01775644388463762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,3072,0.007615110940403408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,3584,0.016195555528004963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,3072,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,2048,0.004201777693298128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,2560,0.00701333334048589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,3072,0.015327110886573792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,2048,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,5120,0.0069057775868309875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,2560,0.013802666631009845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,1536,0.003751111113362842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,1536,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,1024,0.003165333428316646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,1536,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,1536,0.01181244436237547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,1024,0.005676444619894028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,768,0.0028035555862718155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,1024,0.006231110956933763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,2560,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,1024,0.010692444112565784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,512,0.0024844445288181305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,768,0.00599644457300504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,768,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,2560,0.0074106670088238185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,256,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,512,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,512,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,2048,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,128,0.002001777705219057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,256,0.005753777921199799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,512,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,256,0.004994666824738185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,64,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,2048,0.013470222552617392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,128,0.005623111294375525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,256,0.009653333160612319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,65536,0.01482399966981676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,64,0.00509688879052798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,128,0.007430222299363878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,51200,0.012958221965365939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,32,0.004948444250557157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,65536,0.048998223410712353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,65536,0.05621244510014852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,16384,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,51200,0.039686222871144615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,51200,0.04443200098143684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,65536,0.1482506725523207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,12288,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,16384,0.01755911111831665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,16384,0.016493333710564505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,768,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,51200,0.11716355217827691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,10240,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,12288,0.014559111661381192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,8192,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,10240,0.012732444538010491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,16384,0.04351377818319532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,10240,0.01276622215906779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,7168,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,12288,0.03427911135885451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,256,32,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,8192,0.011061333119869232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,8192,0.011556444068749746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,6144,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,7168,0.010434666441546546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,7168,0.010355555348926121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,8192,0.025702221526039973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,5120,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,6144,0.009749333063761393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,6144,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,4096,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,7168,0.02366222275627984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,5120,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,6144,0.021653332644038733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,5120,0.008626666333940294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,3584,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,4096,0.008386666576067606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,4096,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,5120,0.018925334016482036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,3072,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,3584,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,3584,0.0073902226156658595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,12288,0.013782222237851886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,4096,0.01719999975628323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,3072,0.007701333198282454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,3072,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,2048,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,2560,0.006666666517655055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,3072,0.015460444821251763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,10240,0.029859556092156306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,1536,0.003563555578390757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,2048,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,2560,0.013804444836245643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,2048,0.0069857777820693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,1024,0.0029893333299292457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,1536,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,1536,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,2048,0.01309866706530253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,768,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,1024,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,1536,0.011754666765530905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,1024,0.011038222246699862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,768,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,2560,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,512,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,3584,0.016480889585283067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,768,0.006071110980378256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,768,0.010339555641015371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,256,0.0021608887861172357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,2560,0.007455999652544658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,512,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,128,0.0019271110908852683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,256,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,256,0.005657777604129579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,512,0.010572444233629437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,64,0.001965333397189776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,128,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,256,0.009312888814343346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,128,0.004595555365085602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,128,32,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,65536,0.011179555621412067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,51200,0.010173333187898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,64,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,1024,0.00630044440428416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,16384,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,128,0.007270221908887227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,12288,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,32,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,65536,0.05481600099139743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,8192,0.005656888915432825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,51200,0.04378844300905863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,7168,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,16384,0.01753155555990007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,12288,0.014427555931939019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,6144,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,10240,0.012970666918489667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,5120,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,4096,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,8192,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,3584,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,7168,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,3072,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,512,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,6144,0.009648000200589498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,2560,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,2048,0.004862222406599256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,5120,0.008624888956546783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,4096,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,1536,0.0034871109657817413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,3584,0.007678222325113084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,1024,0.0028488888508743713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,768,0.0025395556456512874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,3072,0.006990222467316522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,2560,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,512,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,256,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,2048,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,128,0.001855111163523462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,1536,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,64,0.001855111163523462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,1024,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,32,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,768,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,512,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,65536,0.011210666762457954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,256,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,64,10240,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,16384,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,128,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,12288,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,10240,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,64,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,32,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,8192,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,7168,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,65536,0.054550220568974815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,51200,0.04384088847372267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,16384,0.017476444443066914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,12288,0.014440889159838358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,10240,0.012792888614866467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,8192,0.011408888631396823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,4096,0.0045422220395671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,7168,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,3584,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,3072,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,5120,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,2560,0.005197333378924264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,4096,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,2048,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,3584,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,1536,0.003559111307064692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,2560,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,1024,0.0028479999552170434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,2048,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,768,0.002842666581273079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,1536,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,51200,0.009692444569534725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,512,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,256,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,1024,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,768,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,128,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,512,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,6144,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,64,0.0018168888572189545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,32,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,256,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,6144,0.009854222337404886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,64,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,32,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,3072,0.007052444749408298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,16384,0.3927431106567383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,16384,0.3368053436279297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,224,32,5120,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,51200,1.2468897501627605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,16384,0.13963822523752847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,51200,1.053984006245931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,12288,0.3016142315334744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,10240,0.25181955761379665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,12288,0.2513982190026177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,51200,0.419448004828559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,10240,0.21398754914601645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,8192,0.20160444577534994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,10240,0.08806488911310832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,12288,0.10549333360460068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,7168,0.17725066343943277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,8192,0.07256533039940728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,6144,0.16995378335316977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,8192,0.17067555586496988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,7168,0.06381244129604764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,5120,0.13309066825442845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,7168,0.14869244893391928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,6144,0.128346668349372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,6144,0.05626844697528415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,4096,0.10689510901769002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,128,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,5120,0.048166222042507596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,5120,0.11278400156233047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,3584,0.09160177575217353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,3072,0.08062044117185804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,2560,0.06661066744062635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,4096,0.04035022192531162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,4096,0.0899511112107171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,3584,0.03586311141649882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,2048,0.054641776614718966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,3584,0.08005333609051175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,3072,0.07021511263317533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,1536,0.04247200157907274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,2560,0.027884443600972492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,2560,0.059559113449520536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,1024,0.030870222383075293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,2048,0.02444711161984338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,768,0.024069334069887798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,2048,0.050256000624762646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,1536,0.01904177831278907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,512,0.01759999990463257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,1024,0.015591111448076038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,1536,0.03962399893336826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,256,0.011385777758227454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,768,0.014109333356221518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,1024,0.030964444080988567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,768,0.0251893334918552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,512,0.025576000412305195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,256,0.011747555600272285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,64,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,128,0.011723555624485016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,32,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,256,0.018952000472280715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,64,0.017889777819315594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,128,0.01816177699300978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,32,0.01791733337773217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,3072,0.03199199835459391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,128,0.009205333060688442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,65536,1.057271957397461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,65536,1.2832533518473308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,51200,0.8318195343017578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,51200,0.3298586739434136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,12288,0.24428354369269478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,65536,0.4137270980411106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,16384,0.11089777946472168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,16384,0.2629324330223931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,10240,0.21511733531951904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,12288,0.08319199747509427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,512,0.013392888837390475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,10240,0.07045511404673259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,8192,0.16683377159966362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,7168,0.1434239943822225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,51200,0.9995049370659722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,10240,0.17557956112755668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,8192,0.13877867327796087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,6144,0.13170844978756374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,7168,0.12035466565026177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,5120,0.10613599750730728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,7168,0.051289776961008705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,6144,0.04486844605869717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,4096,0.08763555685679118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,6144,0.10219644175635444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,5120,0.08874133560392593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,3584,0.07623200284110175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,4096,0.07210399707158406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,4096,0.0331857767370012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,5120,0.03840355409516229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,3072,0.06850755876964994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,12288,0.20606044928232828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,3584,0.06303377946217854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,3584,0.02955289019478692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,2560,0.057872891426086426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,2048,0.04679022232691447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,3072,0.026480888326962788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,3072,0.05455555518468221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,2560,0.023740443918440077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,1536,0.036327110396491155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,16384,0.32162311342027455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,2560,0.04804355568355984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,1024,0.02571022179391649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,2048,0.040440890524122454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,1536,0.01648888819747501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,768,0.01984355515903897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,1536,0.032702220810784235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,1024,0.025294222765498694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,1024,0.013782222237851886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,768,0.012699555191728802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,256,0.009633777870072259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,768,0.020781333247820537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,512,0.011508444117175208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,128,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,512,0.020622221959961783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,256,0.010395555860466426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,64,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,256,0.016047999262809753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,32,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,128,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,128,0.0151573336786694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,64,0.0151582227812873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,32,0.014787554740905762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,8192,0.05727022224002414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,65536,0.3827395439147949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,2048,0.019959110352728102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,51200,0.29684800571865505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,16384,0.09771377510494655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,65536,0.35195289717780215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,512,0.014096000128322177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,65536,0.14279645019107395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,12288,0.07495555612776014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,51200,0.28142134348551434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,16384,0.09086844656202529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,10240,0.06494310829374525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,16384,0.04029600156678094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,8192,0.052496890226999916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,51200,0.11254310607910156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,12288,0.0322364436255561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,7168,0.04615288972854614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,12288,0.06877688566843669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,10240,0.027890665663613215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,6144,0.0396382212638855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,10240,0.05851022402445475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,8192,0.02351555559370253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,5120,0.03407466742727492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,8192,0.04904266529613071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,4096,0.027466666367318895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,6144,0.019038221902317472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,7168,0.04336622357368469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,5120,0.016840888394249808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,7168,0.021340444684028625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,6144,0.03785066803296407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,3072,0.022299556268586054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,4096,0.01499555508295695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,2560,0.0185253338681327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,3584,0.013721778160995908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,3072,0.01297511077589459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,4096,0.027560888065232172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,5120,0.032827556133270264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,2048,0.014887111054526435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,1536,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,2560,0.01204177737236023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,2048,0.01108977778090371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,2560,0.01921688848071628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,1024,0.009372444616423713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,2048,0.016765332884258695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,1536,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,768,0.007681777907742395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,1536,0.01408622165520986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,512,0.006220444622966979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,1024,0.007868444754017724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,1024,0.011689778003427716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,768,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,256,0.004304000073009067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,768,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,512,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,256,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,512,0.010067555639478896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,64,0.003306666596068276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,256,0.008620444271299574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,3584,0.025332444243960913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,32,0.0032506665835777917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,128,0.00811111099190182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,64,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,32,0.008054222497675154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,65536,0.2960888809627957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,51200,0.2269653214348687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,65536,0.28496265411376953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,65536,0.1268355581495497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,16384,0.08125778039296468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,128,0.0038319999972979226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,51200,0.10084266795052423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,51200,0.22488710615370008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,12288,0.0627875526746114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,16384,0.07620355818006727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,16384,0.0376835564772288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,128,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,10240,0.04940711127387153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,8192,0.04209688968128628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,12288,0.029674665795432195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,12288,0.05804710918002658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,10240,0.025648888614442613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,3584,0.024804444776640996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,10240,0.04965777860747444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,3072,0.021839110387696162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,6144,0.033124443557527326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,7168,0.019987555013762582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,5120,0.02773333258099026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,7168,0.03674044542842441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,6144,0.017960889471901786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,4096,0.023471110396915015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,6144,0.03224000003602769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,5120,0.016188444362746347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,3584,0.02051466703414917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,4096,0.01441511180665758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,5120,0.02784622377819485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,3072,0.01792533364560869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,4096,0.023238221804300945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,3584,0.013330666555298699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,2560,0.014920888675583733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,3072,0.012610666453838348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,3584,0.02125688890616099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,2048,0.012845333251688214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,3072,0.019144000278578866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,2560,0.01145155562294854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,1536,0.010676444404655032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,2560,0.016961778203646343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,7168,0.03773599863052368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,2048,0.01458755632241567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,1024,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,1536,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,8192,0.02166044380929735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,1536,0.012773333324326409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,8192,0.04145688811937968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,1024,0.007699555820888943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,512,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,1024,0.01034311122364468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,768,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,256,0.004504888835880491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,512,0.007002666592597961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,768,0.008804444637563493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,128,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,512,0.00851466672288047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,256,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,256,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,32,0.0032133333798911837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,128,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,128,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,64,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,32,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,51200,0.21500355667538115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,2048,0.010051555931568146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,65536,0.24269154336717394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,65536,0.12318933010101318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,16384,0.07913600073920356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,12288,0.062295112344953746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,768,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,16384,0.03628888726234436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,51200,0.09877510865529378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,10240,0.051185776789983116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,16384,0.06492355797025892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,12288,0.02830044428507487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,12288,0.049864888191223145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,8192,0.04222222169240316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,7168,0.03611110978656345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,10240,0.024711110525661047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,8192,0.021313778228229944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,10240,0.04299555553330315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,6144,0.03322311242421468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,8192,0.03603733248180813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,7168,0.01928711103068458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,5120,0.02648266653219859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,7168,0.031558222240871854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,6144,0.01737866633468204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,65536,0.2545928955078125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,4096,0.022340445054901972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,3584,0.019839111301634047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,6144,0.0281333327293396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,5120,0.0239982224173016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,3072,0.01757600075668759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,3584,0.013095999757448832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,4096,0.014203555054134794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,4096,0.020637333393096924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,2560,0.015015999476114908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,3584,0.0185199992524253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,51200,0.19137422243754068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,3072,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,2048,0.013220444321632385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,2560,0.010514666636784872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,3072,0.016873778568373788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,1536,0.009988444546858469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,2560,0.015153777268197803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,1024,0.008100444244013892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,2048,0.00886311133702596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,1536,0.00829777783817715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,64,0.0031822222388452957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,768,0.006523555351628198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,1536,0.011528000235557556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,512,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,1024,0.009069333473841349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,768,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,256,0.003855111284388436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,768,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,512,0.006896889044178857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,128,0.0032186667538351486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,512,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,5120,0.015777778294351365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,256,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,256,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,64,0.0031999999450312722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,128,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,128,0.006800888727108638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,64,0.006993778049945831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,32,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,65536,0.220015115208096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,51200,0.1707271072599623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,65536,0.1985679997338189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,65536,0.0763253304693434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,2048,0.013447999954223633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,16384,0.06105155415005154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,1024,0.0074142225914531285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,12288,0.04698666599061754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,51200,0.15483821762932673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,16384,0.0239991115199195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,10240,0.037464890215131975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,51200,0.06121777825885349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,16384,0.05368622144063314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,8192,0.03228977653715346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,10240,0.016792888442675274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,12288,0.04141511188613044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,8192,0.01496355566713545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,7168,0.027444443768925134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,6144,0.023991111252042983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,8192,0.030066665675905015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,32,0.0035111109415690103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,5120,0.02090666691462199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,7168,0.013850666582584381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,6144,0.01275466630856196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,7168,0.026504000027974445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,4096,0.017245334055688646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,5120,0.011695110963450538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,6144,0.023360000716315374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,3584,0.014752000570297241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,5120,0.020417778028382193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,4096,0.010770666930410596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,3584,0.009648000200589498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,4096,0.017709333035680983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,3584,0.016214221715927124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,2560,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,3072,0.008616888688670265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,2048,0.010016000105275048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,3072,0.014834667245546976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,1536,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,2560,0.007879999776681265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,12288,0.01924977699915568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,2048,0.0075733330514695905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,2560,0.013452444639470843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,1024,0.006690666493442323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,768,0.005169777820507686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,2048,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,1536,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,512,0.004111111164093018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,1024,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,768,0.00655644428398874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,1024,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,256,0.003208000005947219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,768,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,512,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,128,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,512,0.007297777467303806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,256,0.005886222339338726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,128,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,32,0.002867555452717675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,128,0.006302222195598815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,3072,0.013455111119482251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,64,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,32,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,65536,0.20581511656443277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,51200,0.16629778014289007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,65536,0.19408888287014434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,65536,0.07236978080537584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,16384,0.05434755484263102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,10240,0.036066667901145086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,1536,0.009741333623727163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,51200,0.15153422620561388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,51200,0.05863822168774075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,12288,0.04283110962973701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,16384,0.02267644471592373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,16384,0.05303466651174757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,12288,0.01852888862291972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,8192,0.03007555670208401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,12288,0.04049244523048401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,10240,0.016521778371598985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,256,0.006575999988449945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,7168,0.026551110876931086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,8192,0.014560000763999091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,6144,0.022967111733224656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,7168,0.013436444103717804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,8192,0.029149333635965984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,5120,0.01919644408755832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,7168,0.0258559998538759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,6144,0.01240711079703437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,4096,0.016216889023780823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,5120,0.011368888947698804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,6144,0.02286311156219906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,3584,0.014354666074117025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,5120,0.01984444426165687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,4096,0.010068444742096795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,3072,0.012989333106411828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,4096,0.017429333594110277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,3584,0.009388444324334463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,2560,0.011010666688283285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,3584,0.01590222285853492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,3072,0.008513777620262569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,10240,0.036121779017978244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,2048,0.009644444617960189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,3072,0.014535999960369535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,2560,0.007715555528799693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,1536,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,64,0.0025928889711697898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,2048,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,2560,0.012842666771676807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,1024,0.006121777825885349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,1536,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,2048,0.011160888605647616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,768,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,512,0.004037333445416557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,1536,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,768,0.006404444575309753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,1024,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,256,0.0032506665835777917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,512,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,128,0.002967111145456632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,512,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,256,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,64,0.0028284444577164124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,256,0.006504000061088138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,128,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,32,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,128,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,64,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,65536,0.17208888795640734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,32,0.0063279999627007385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,51200,0.14051822821299234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,65536,0.07032444741990831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,16384,0.04863377743297153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,10240,0.03474666674931844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,51200,0.056458665264977344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,51200,0.12201778093973796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,16384,0.021715554926130507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,16384,0.04330933425161573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,10240,0.03285333183076646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,768,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,12288,0.017850667238235474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,8192,0.030630220969518025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,10240,0.01607822212908003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,10240,0.02877955635388692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,7168,0.025489777326583862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,8192,0.014121777481502958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,6144,0.023571555813153584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,8192,0.02404888967672984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,7168,0.013199999928474426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,5120,0.01927911076280806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,7168,0.021692444880803425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,6144,0.012145777543385824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,6144,0.019281778070661757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,4096,0.0144213338692983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,5120,0.010952889091438718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,65536,0.15470400121476915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,3584,0.013270222478442721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,5120,0.016890666551060148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,4096,0.009307555854320526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,3072,0.01145333300034205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,4096,0.014896000425020853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,12288,0.038344889879226685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,2560,0.010076444182131026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,1024,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,3072,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,3584,0.013759999639458127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,2560,0.007709333466158972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,3072,0.0127368892232577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,1536,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,2560,0.011428444749779172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,1024,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,2048,0.00965244405799442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,1536,0.008356444537639618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,1024,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,768,0.004853333450025982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,1024,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,768,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,512,0.003830222205983268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,768,0.006796444455782573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,512,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,256,0.003136000078585413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,512,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,2048,0.009120889008045197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,128,0.0025235555238193935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,256,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,256,0.006144888699054718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,12288,0.03348533312479655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,64,0.002465777720014254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,32,0.002587555597225825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,128,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,1536,0.0069004446268081665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,128,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,65536,0.13505866792466906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,64,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,32,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,51200,0.11067022217644586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,65536,0.06764533122380574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,16384,0.04568799999025133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,65536,0.15211733182271323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,12288,0.03779110974735684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,51200,0.05435555511050754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,51200,0.11885955598619248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,16384,0.021015110943052504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,10240,0.03312355611059401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,16384,0.041929778125551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,12288,0.017316444052590262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,8192,0.028320001231299505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,3584,0.00869599978129069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,12288,0.032080888748168945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,7168,0.025806221697065566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,10240,0.015569777952300178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,10240,0.027460444304678176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,6144,0.022132444712850783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,8192,0.013803555733627744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,5120,0.018612444400787354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,7168,0.012776888906955719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,8192,0.023457777169015672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,2048,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,6144,0.011723555624485016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,4096,0.016386666231685214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,3584,0.01367111090156767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,5120,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,4096,0.008849778109126622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,5120,0.016807110773192514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,3072,0.010767999622556897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,3584,0.008363555702898238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,4096,0.014599111345079211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,2560,0.009327111144860586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,3584,0.013472889032628802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,3072,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,3072,0.01239822225438224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,2560,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,1536,0.006909333169460297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,2560,0.01071999967098236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,1024,0.005114666703674528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,2048,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,1536,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,2048,0.009343110852771336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,768,0.0042071110672420925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,1536,0.00832266691658232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,1024,0.0063573333124319715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,1024,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,768,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,256,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,768,0.006696888970004187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,512,0.006041777630647023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,7168,0.021198223034540813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,512,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,256,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,6144,0.01886488828394148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,64,0.002540444541308615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,256,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,32,0.0026417778184016547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,128,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,128,0.0058782220714622075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,65536,0.1181288825141059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,2048,0.00812711069981257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,32,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,51200,0.09528177976608276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,65536,0.0653288894229465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,16384,0.03751822312672933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,65536,0.11143733395470513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,12288,0.02882044514020284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,51200,0.05180622140566508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,51200,0.08860711256663005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,16384,0.03201422095298767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,10240,0.024241778585645888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,12288,0.02478933334350586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,8192,0.019863999552196927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,10240,0.015095111396577625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,10240,0.021706667211320665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,128,0.0028488888508743713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,7168,0.01759466694460975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,8192,0.013155555559529198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,8192,0.0185244447655148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,7168,0.012200888660218982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,6144,0.015428443749745687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,7168,0.016893333858913846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,5120,0.01348088930050532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,64,0.005970666805903117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,6144,0.010806221928861407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,6144,0.01532799998919169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,4096,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,5120,0.009666666388511658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,5120,0.013703111145231458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,3584,0.012329777909649743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,4096,0.008649778034951951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,16384,0.020233778489960563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,3072,0.011048888994587792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,4096,0.012178666889667511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,3584,0.00831288927131229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,12288,0.01686044368478987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,2560,0.00962577760219574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,2048,0.008264888491895463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,3584,0.010944888823562197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,3072,0.007932444413503012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,2560,0.007618666523032718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,3072,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,2048,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,1024,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,2048,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,1536,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,768,0.004615111069546806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,1536,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,1024,0.0064186664919058485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,512,0.0038515557017591265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,1024,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,768,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,256,0.0028764444092909494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,512,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,128,0.0025315555847353404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,512,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,256,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,64,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,128,0.0058124446206622655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,32,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,128,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,64,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,65536,0.10420266787211101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,1536,0.007072000039948358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,32,0.005877333382765452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,2560,0.008843555218643611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,51200,0.0746506651242574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,65536,0.06369066900677152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,16384,0.03606222073237101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,65536,0.11265067259470622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,51200,0.05066666669315762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,12288,0.028781334559122723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,51200,0.08746311399671768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,16384,0.019946666227446664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,10240,0.023103111320071753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,16384,0.03124888737996419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,12288,0.016591111818949383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,8192,0.019181332654423185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,12288,0.024803555674023096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,10240,0.015034667319721647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,7168,0.01686755485004849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,8192,0.012865777644846173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,10240,0.02162666618824005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,6144,0.014907555447684394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,8192,0.01841600073708428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,7168,0.011694221860832639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,5120,0.013015111287434896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,4096,0.011691555380821228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,7168,0.01687022215790219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,512,0.003555555724435382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,6144,0.01035377797153261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,5120,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,6144,0.015202666322390238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,5120,0.0136835558546914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,3072,0.008737777670224508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,4096,0.008647110727098253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,2560,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,3584,0.008279999925030602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,4096,0.012039110892348819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,2048,0.0068693334857622785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,3072,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,2560,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,3072,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,1536,0.006335111127959357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,2560,0.008732444710201686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,2048,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,1024,0.004549333204825719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,2048,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,1536,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,768,0.0038382220599386427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,1024,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,1536,0.007258666886223688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,512,0.003195555466744635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,768,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,1024,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,256,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,768,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,512,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,128,0.002469333302643564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,512,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,256,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,64,0.0021928888228204516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,256,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,128,0.005799110978841782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,3584,0.009766221874290043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,32,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,128,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,65536,0.0877457790904575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,64,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,51200,0.06687288814120822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,3584,0.010763555765151978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,16384,0.027052443888452318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,65536,0.06263022290335761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,65536,0.10567822058995564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,12288,0.022226666410764057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,51200,0.0500311123000251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,16384,0.019687111179033916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,51200,0.08346311251322429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,10240,0.02239911092652215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,16384,0.030160887373818293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,12288,0.01639111174477471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,8192,0.017482666505707633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,12288,0.0241333329015308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,10240,0.014775110615624322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,7168,0.014531556102964612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,10240,0.020993777447276648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,8192,0.012497777740160624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,6144,0.012842666771676807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,8192,0.01806400054030948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,5120,0.011749332977665795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,7168,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,6144,0.009995555712117089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,7168,0.016520889268981088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,4096,0.00976533359951443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,6144,0.014846222268210517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,5120,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,3584,0.009886222581068674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,5120,0.013191999660597907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,4096,0.008631111019187504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,3072,0.008886222210195329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,4096,0.011152889165613385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,3584,0.008347555167145198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,32,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,2560,0.007415999968846639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,2048,0.006221333311663733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,3072,0.00795999997191959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,3584,0.010135110881593492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,1536,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,3072,0.009365333451165093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,2048,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,1024,0.003920888735188378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,1536,0.006884444504976273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,2048,0.007991111112965478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,768,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,768,0.0035004446076022256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,1536,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,512,0.0030871110243929755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,1024,0.006302222195598815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,768,0.006224000205596288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,1024,0.006670222100284364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,256,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,768,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,512,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,128,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,256,0.0058764442801475525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,512,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,64,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,32,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,128,0.005687110953860813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,256,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,128,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,65536,0.07979910903506808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,51200,0.06526399983300103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,32,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,65536,0.06165066692564222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,16384,0.0255395554833942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,65536,0.10361066791746353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,2560,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,51200,0.04895288745562235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,51200,0.08200444115532769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,2560,0.008784000244405534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,16384,0.01961777773168352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,10240,0.0177502218219969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,16384,0.02968177861637539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,12288,0.016040888097551134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,10240,0.014096000128322177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,12288,0.02347022294998169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,7168,0.012690666649076672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,8192,0.011539555258221097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,10240,0.020658666888872784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,6144,0.01130311108297772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,8192,0.01771022213829888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,7168,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,5120,0.009933333430025313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,7168,0.016149333781666227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,4096,0.009875555833180746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,6144,0.014375110467274984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,5120,0.009401777552233802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,5120,0.012416000167528788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,3584,0.009206222163306342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,4096,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,64,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,3072,0.009187555147541894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,4096,0.01072799993885888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,3584,0.008359111017651027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,2560,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,12288,0.020536889632542927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,3072,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,2048,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,3072,0.009320889082219865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,8192,0.018059555027219985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,1536,0.006179555422729916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,2560,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,2048,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,1024,0.0041697778635554844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,2048,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,1536,0.006653333289755716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,768,0.0038524443904558816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,1536,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,1024,0.0063386667105886675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,6144,0.010001777774757808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,512,0.003195555466744635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,1024,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,256,0.002809777855873108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,768,0.006142222219043308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,512,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,768,0.006169777777459886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,128,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,512,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,256,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,64,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,256,0.005384888913896348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,32,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,128,0.005399111244413588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,65536,0.06635289059744941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,3584,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,64,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,32,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,51200,0.05072177780999077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,2560,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,16384,0.021147555775112573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,65536,0.06490666336483426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,12288,0.01715022159947289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,51200,0.047268443637424044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,51200,0.05078666739993625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,16384,0.019208888212839764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,10240,0.01696977847152286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,16384,0.019941333267423842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,12288,0.015286222100257874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,8192,0.013112000293201871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,12288,0.016241777274343703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,10240,0.01351466692156262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,7168,0.01239111108912362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,10240,0.01464355488618215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,8192,0.011402666568756104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,6144,0.010464000205198923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,8192,0.01276533305644989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,7168,0.010753778119881948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,6144,0.010079999764760336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,4096,0.008054222497675154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,128,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,7168,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,5120,0.009301332963837517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,3584,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,5120,0.009320889082219865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,3072,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,4096,0.008655110994974772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,3584,0.008309333688682979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,4096,0.008307555483447181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,2560,0.005798222290145026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,3584,0.008016000191370646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,3072,0.007815111014578078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,2048,0.004966222163703707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,3072,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,2560,0.00755466686354743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,1536,0.00423911131090588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,2048,0.007240888973077138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,1024,0.003559111307064692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,2560,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,2048,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,1536,0.006755555669466655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,768,0.003195555466744635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,1536,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,1024,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,512,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,768,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,1024,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,256,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,768,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,512,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,5120,0.009317333499590555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,128,0.002303999952144093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,64,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,256,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,512,0.005362666729423735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,32,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,128,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,6144,0.010067555639478896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,128,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,65536,0.04660800099372864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,64,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,51200,0.0381644434399075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,32,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,16384,0.016113777955373127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,65536,0.05834844377305773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,65536,0.061908443768819175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,12288,0.015795555379655626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,51200,0.046446220742331616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,51200,0.04897955391142103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,10240,0.013447999954223633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,16384,0.01962844365172916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,12288,0.014129777749379476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,8192,0.012291555603345236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,12288,0.015807999504937064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,10240,0.013038222160604266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,7168,0.011680888632933298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,10240,0.014197332991494073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,8192,0.011358222199810876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,6144,0.01033688916100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,8192,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,7168,0.010665777656767102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,5120,0.009026666482289633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,6144,0.00998044427898195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,256,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,4096,0.007985778152942657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,6144,0.0099982221921285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,5120,0.00924533357222875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,3584,0.007247111035717859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,4096,0.008519111408127679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,5120,0.009321778184837764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,3072,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,65536,0.0600186652607388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,4096,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,2560,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,3584,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,16384,0.018412444326612685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,3584,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,3072,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,2048,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,2560,0.007463999920421177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,1536,0.004863111095296012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,3072,0.007612444460391998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,1024,0.0037866665257347953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,2560,0.007130666739410824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,2048,0.007033777733643849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,768,0.0032239999208185407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,1536,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,2048,0.00656444455186526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,512,0.0028533333291610083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,1024,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,1536,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,1024,0.0058755555914507965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,256,0.002595555658141772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,768,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,256,0.005635555419656966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,128,0.0021520000365045336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,512,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,7168,0.011006222003036074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,64,0.0021377777059872947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,32,0.0021280000607172647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,256,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,128,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,65536,0.03491200009981791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,128,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,51200,0.028541333145565454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,64,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,65536,0.057329780525631376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,32,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,16384,0.014116444521480136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,12288,0.011976888610257043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,65536,0.05942400296529134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,51200,0.045851555135515004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,10240,0.011983999775515662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,51200,0.04691733254326714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,16384,0.016694222887357075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,8192,0.01071111112833023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,12288,0.014106666048367819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,16384,0.018852444158660043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,7168,0.00997155573632982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,10240,0.012720888687504662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,12288,0.014932443698247274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,6144,0.009156444834338294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,7168,0.010654222634103564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,768,0.006124444305896759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,5120,0.010982222027248807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,7168,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,512,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,6144,0.009752888646390703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,5120,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,5120,0.009012444151772393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,3584,0.007135999699433644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,4096,0.008407110969225565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,4096,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,3072,0.006524444454246097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,3584,0.008204444415039485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,3584,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,3072,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,2560,0.00583022211988767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,3072,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,2560,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,8192,0.01108088923825158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,2048,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,2560,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,8192,0.01145333300034205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,1536,0.004231111043029361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,2048,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,6144,0.009738666315873465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,1024,0.0034888887570963967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,1536,0.006790222393141852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,4096,0.008153777983453538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,768,0.003559111307064692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,2048,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,1536,0.006215111249023014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,1024,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,768,0.006167111297448476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,256,0.0025351111673646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,768,0.005628444254398346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,512,0.0058808889653947614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,128,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,512,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,256,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,64,0.0022044444663657085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,256,0.005294222384691238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,32,0.0021608887861172357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,128,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,65536,0.028386665715111628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,64,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,32,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,10240,0.013408888545301227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,51200,0.02372177773051792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,65536,0.057052446736229785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,65536,0.057894223266177706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,51200,0.0457911127143436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,51200,0.046022223101721875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,512,0.002826666666401757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,16384,0.016573333077960543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,16384,0.01833511061138577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,1024,0.0058319999112023254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,10240,0.010827555424637265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,12288,0.014119999276267158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,8192,0.010029333333174387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,12288,0.01479200025399526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,10240,0.01275111072593265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,7168,0.009394666386975182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,128,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,10240,0.013129777378506131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,8192,0.011203555597199334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,6144,0.008436444732877944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,5120,0.007763555480374231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,7168,0.010657777388890585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,6144,0.009705777797434065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,4096,0.008368888662921058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,6144,0.009679999616410997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,5120,0.009119110802809397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,3584,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,5120,0.008946666287051307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,4096,0.008393777741326226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,3072,0.005877333382765452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,3584,0.008033778104517195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,3584,0.007768889268239339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,12288,0.01260444439119763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,2560,0.005187555733654234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,3072,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,2048,0.004512000001139111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,3072,0.00739644467830658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,2560,0.007342221836249034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,1536,0.004159111115667555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,2560,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,2048,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,1024,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,1536,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,1536,0.005998222364319696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,768,0.003169777699642711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,8192,0.011393778026103973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,1024,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,512,0.0028328889360030494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,1024,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,768,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,256,0.002504000026318762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,512,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,512,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,16384,0.014067555467287699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,256,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,128,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,4096,0.008018666671382057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,64,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,256,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,768,32,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,128,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,65536,0.024290665984153748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,128,0.004955555415815777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,64,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,51200,0.019839111301634047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,32,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,16384,0.010650667051474253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,2048,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,65536,0.05657511287265354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,51200,0.044212443961037524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,12288,0.009123555488056606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,51200,0.045399109522501625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,16384,0.016543111867374845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,10240,0.012219555675983429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,16384,0.01721955504682329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,12288,0.013975999421543546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,8192,0.010212444596820408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,12288,0.014517333772447375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,10240,0.012814222110642327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,7168,0.00925511121749878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,10240,0.012852444416946836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,8192,0.01108266661564509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,6144,0.008608000146018134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,8192,0.011372444530328115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,7168,0.010401777923107147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,5120,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,7168,0.010703110860453712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,6144,0.009711110757456886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,4096,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,6144,0.009710222482681274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,5120,0.009085333181752099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,3584,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,5120,0.008977777428097194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,65536,0.05678755707210965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,4096,0.008336000144481659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,3072,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,4096,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,3584,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,2560,0.004947555561860402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,3584,0.007696000238259633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,3072,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,2048,0.00453955555955569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,3072,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,1536,0.004234666625658671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,2560,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,2048,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,7168,0.010732444624106089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,1024,0.0032097777972618737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,1536,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,768,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,768,0.003330666571855545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,1536,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,1024,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,768,0.006227555374304454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,1024,0.005728000154097875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,256,0.002287111141615444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,768,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,512,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,128,0.002151111140847206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,512,0.005290666802061929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,256,0.0058853332367208265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,64,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,128,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,256,0.00518844442235099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,32,0.0021537778278191886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,65536,0.018208000395033095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,128,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,64,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,51200,0.01572088897228241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,65536,0.04978222317165799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,2560,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,16384,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,2048,0.007118222614129384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,65536,0.05550488829612732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,12288,0.008586666650242275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,51200,0.03972622089915805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,10240,0.007756444315115611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,16384,0.01664266652531094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,51200,0.044807109567854136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,12288,0.014093332820468478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,512,512,0.0025635556214385563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,10240,0.012800888882742988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,12288,0.014110222458839417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,7168,0.008584000170230865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,10240,0.012727999852763282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,8192,0.01107377807299296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,6144,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,8192,0.011371555427710215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,7168,0.010660444696744284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,7168,0.01034577770365609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,6144,0.009409777820110321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,4096,0.006852444261312485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,5120,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,5120,0.00889066689544254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,3584,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,4096,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,32,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,4096,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,3072,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,3584,0.007962666451931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,2560,0.004871110949251386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,3584,0.007672888537247975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,3072,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,2048,0.004463110946946674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,16384,0.017197334104114108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,2560,0.0069120000633928515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,1536,0.003661333272854487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,2048,0.006991110742092133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,2048,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,5120,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,1024,0.003157333367400699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,6144,0.009667555491129557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,1536,0.006015111174848344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,768,0.002797333316670524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,1024,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,1024,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,512,0.0024702221983008915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,768,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,768,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,256,0.0022008888837363985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,512,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,512,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,128,0.001957333336273829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,256,0.005628444254398346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,256,0.005052444421582752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,64,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,8192,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,256,32,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,128,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,65536,0.01553600033124288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,128,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,64,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,51200,0.016580444243219163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,65536,0.04926133486959669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,1536,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,65536,0.05340800020429823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,16384,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,51200,0.039721777041753135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,12288,0.010674667027261523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,51200,0.042816887299219765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,16384,0.016521778371598985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,10240,0.008347555167145198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,12288,0.013756444056828817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,8192,0.008336000144481659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,12288,0.014442667365074158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,10240,0.012440000143316058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,7168,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,10240,0.012712889247470431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,8192,0.011039111349317761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,8192,0.011044444309340583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,2560,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,7168,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,5120,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,7168,0.010256888965765635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,6144,0.009661333428488838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,4096,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,6144,0.009418666362762451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,5120,0.008991111483838823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,32,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,3584,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,5120,0.008609777523411645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,4096,0.008375999828179678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,3072,0.00553955551650789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,4096,0.007687110867765214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,3584,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,2560,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,3584,0.007301333049933116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,3072,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,16384,0.016885333591037326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,2560,0.007343110938866933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,3072,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,1536,0.0035662220584021676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,2560,0.006660444455014334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,1024,0.0030257778449190986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,2048,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,1536,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,2048,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,768,0.0028088889602157804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,1024,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,512,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,6144,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,3072,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,768,0.0060159998635451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,256,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,768,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,512,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,256,0.005629333357016246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,512,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,64,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,256,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,32,0.0018275556051068837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,128,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,65536,0.02421333392461141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,128,0.004876444323195351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,51200,0.020578665865792167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,64,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,16384,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,32,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,2048,0.0041759999261962045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,12288,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,10240,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,65536,0.05269244313240051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,8192,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,51200,0.042451557185914784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,7168,0.006298666612969504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,16384,0.01684799955950843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,6144,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,12288,0.014077333940400017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,10240,0.012488000094890594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,1536,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,5120,0.006341333190600078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,8192,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,4096,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,1024,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,3584,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,3072,0.0053839998112784485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,6144,0.009328000247478485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,2560,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,128,128,0.0018195555441909367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,5120,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,2048,0.004200889004601372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,4096,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,1536,0.003491555651028951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,3584,0.007342221836249034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,1024,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,768,0.0025822222232818604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,3072,0.006994666324721442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,512,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,2560,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,256,0.0021475555582178966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,2048,0.006331555545330048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,1536,0.005975111077229182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,1024,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,64,0.0018195555441909367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,512,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,32,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,256,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,65536,0.02608888844648997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,128,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,51200,0.022335110439194575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,64,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,16384,0.012436444560686747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,32,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,12288,0.009890666438473595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,65536,0.05269599954287211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,7168,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,10240,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,8192,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,16384,0.016920889417330425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,7168,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,12288,0.013975999421543546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,6144,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,10240,0.012735111018021902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,8192,0.011032889286677042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,7168,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,4096,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,6144,0.009322666459613377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,64,128,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,3584,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,3072,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,768,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,5120,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,2560,0.004855999930037392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,2048,0.0042239998777707415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,4096,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,1536,0.003535110917356279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,3584,0.007676444119877285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,1024,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,3072,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,2560,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,768,0.002663111107216941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,512,0.002492444382773505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,2048,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,256,0.0021093332519133887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,1536,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,128,0.0018719999740521114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,51200,0.042465776205062866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,64,0.001820444439848264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,32,0.0017964444640609955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,512,0.004944889081848992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,256,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,128,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,64,0.004575110971927643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,193,32,5120,0.00647555540005366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,32,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,16384,0.3926977846357558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,16384,0.277237335840861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,1024,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,51200,0.8564657635158963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,768,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,51200,1.2317626741197374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,16384,0.1394719945059882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,12288,0.30383645163642037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,12288,0.20494488875071207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,10240,0.24341244167751738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,51200,0.41464000278049046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,10240,0.1711031066046821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,8192,0.20591555701361763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,12288,0.10440888669755723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,10240,0.08854577938715617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,8192,0.140538665983412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,7168,0.17879288726382783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,8192,0.07195466756820679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,12288,0.23444623417324492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,16384,0.3131493462456597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,7168,0.12062044938405354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,7168,0.06377155250973172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,6144,0.15288266870710585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,10240,0.19696888658735487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,6144,0.10999377568562825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,6144,0.05584444602330526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,8192,0.15717155403561062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,7168,0.13978755474090576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,5120,0.04742933313051859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,5120,0.08860266870922512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,4096,0.10291822089089288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,6144,0.1209146711561415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,3584,0.09465066591898601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,4096,0.040307554933759905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,4096,0.07538488838407728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,3584,0.03570933474434747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,3584,0.06495910882949829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,3072,0.07823733488718669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,5120,0.12492444780137803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,3072,0.05585600270165337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,3584,0.0750782224867079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,2560,0.06864977545208402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,3072,0.03198666705025567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,3072,0.06584178076850043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,2560,0.04893689023123848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,2048,0.05265688896179199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,2560,0.02812088860405816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,5120,0.10176799694697063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,2560,0.05760266383488973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,2048,0.04159199860360887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,4096,0.08534044689602321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,2048,0.05035733183224996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,1024,0.02996088729964362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,1536,0.01906488835811615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,1536,0.042430222034454346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,768,0.02294755478700002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,1024,0.015779554843902588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,1024,0.025920000341203477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,1024,0.03478577733039856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,768,0.014205333259370593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,768,0.021016000045670405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,51200,0.9450986650254992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,512,0.016660445266299777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,256,0.01033333357837465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,768,0.029481778542200725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,512,0.01531733406914605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,512,0.017683555682500202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,2048,0.024392889605628118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,128,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,256,0.013823111024167804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,512,0.025272889269722834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,1536,0.03289066751797994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,64,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,128,0.012433778080675336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,256,0.021013332737816706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,128,0.017197334104114108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,128,0.01479555500878228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,64,0.014800889624489678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,32,0.015503111812803479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,1536,0.042437334855397545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,256,0.01575288838810391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,65536,1.0960124333699544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,32,0.0075839997993575195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,65536,0.8855102327134875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,16384,0.2736373212602404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,51200,0.6827751265631782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,16384,0.21987999810112846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,65536,0.41562843322753906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,12288,0.2087786727481418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,16384,0.10976177453994751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,12288,0.1843057738410102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,51200,0.851836469438341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,10240,0.20136800077226427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,12288,0.08306044340133667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,51200,0.3290657732221815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,10240,0.13694843981001112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,10240,0.06996799839867486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,8192,0.16995644569396973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,16384,0.25191466013590497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,7168,0.14378755622439915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,8192,0.11328355471293132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,8192,0.057542224725087486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,12288,0.1881937715742323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,10240,0.15869688987731934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,6144,0.12228000164031982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,7168,0.05146844519509209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,7168,0.09874133268992107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,8192,0.1282337771521674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,5120,0.10378666718800862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,6144,0.04496533340877957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,5120,0.03850933247142368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,6144,0.0983733336130778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,7168,0.11297510729895698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,4096,0.0893697804874844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,5120,0.07117688655853271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,5120,0.08452444606357151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,3584,0.0737875567542182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,4096,0.03295288814438714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,4096,0.058719999260372586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,3072,0.06586933135986328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,3584,0.029345777299669053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,3584,0.052058666944503784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,4096,0.07026400168736775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,3072,0.046075556013319224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,6144,0.0866853329870436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,3072,0.026713778575261433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,3584,0.06357866525650024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,2560,0.054083555936813354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,2048,0.04407199886110094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,3072,0.05618221892250908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,2560,0.023758222659428913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,2560,0.03983733389112685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,1536,0.03479289015134176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,2048,0.03401777810520596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,51200,0.7955119874742297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,2560,0.05106400118933784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,1024,0.025460445218616064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,2048,0.04298933347066244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,1536,0.02736888825893402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,1536,0.020169778002632987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,768,0.021387555532985266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,1024,0.013901333014170328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,1024,0.0215013325214386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,65536,1.010968844095866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,512,0.01670222149954902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,1024,0.029888000753190782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,768,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,512,0.011716444459226398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,256,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,512,0.018174222773975797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,512,0.022488888767030504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,2048,0.020070221689012315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,256,0.01348888874053955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,256,0.01036800030204985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,256,0.018385777870814007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,128,0.01036177741156684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,64,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,128,0.012513777448071374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,32,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,128,0.014966222974989148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,1536,0.037295109695858426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,64,0.012644444074895648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,32,0.013401777380042605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,768,0.02097333305411869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,65536,0.3809200127919515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,768,0.02646844420168135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,51200,0.2990470992194282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,65536,0.3069350984361437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,16384,0.09701422188017104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,65536,0.14220178127288818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,51200,0.23808799849616158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,128,0.007000889215204451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,12288,0.07526133457819621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,51200,0.11310933695899116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,16384,0.04059377643797133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,16384,0.08049511247211032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,12288,0.06100800302293566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,10240,0.06332711378733318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,12288,0.03211911188231574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,8192,0.05423466695679558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,16384,0.08843466970655654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,10240,0.05234488844871521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,10240,0.027847111225128174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,12288,0.06952799691094293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,7168,0.04693688948949178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,8192,0.04411200020048353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,8192,0.02345688806639777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,10240,0.05948888593249851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,6144,0.039750221702787615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,7168,0.02130488885773553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,51200,0.26049688127305776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,6144,0.03386133246951633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,8192,0.0509786672062344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,6144,0.0188426673412323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,65536,0.3270435598161486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,5120,0.01668177710639106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,6144,0.0418817765182919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,4096,0.02881511052449544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,5120,0.029288000530666772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,3584,0.025441777375009324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,4096,0.015443555182880826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,5120,0.037032001548343234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,4096,0.024737776981459722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,3072,0.022575111852751836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,3584,0.022169777088695105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,3584,0.013879111243618859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,4096,0.0324853327539232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,2560,0.018495111001862418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,3072,0.012870222330093384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,3072,0.01959466603067186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,3584,0.030155556069480047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,2048,0.015633778439627755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,2560,0.012059555285506778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,3072,0.026886221435334947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,7168,0.0383840004603068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,7168,0.046859555774264865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,1536,0.012376888758606382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,2048,0.01514933341079288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,2048,0.010955555571450127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,2560,0.024333332975705464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,1024,0.009820444716347588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,5120,0.0346666673819224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,1536,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,1536,0.012880889077981314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,768,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,1024,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,1024,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,1536,0.019278221660190158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,512,0.006230222268237009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,1024,0.016759999924235873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,768,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,768,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,256,0.004392000122202767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,512,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,512,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,768,0.014561777313550314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,128,0.0035502223504914176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,256,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,512,0.012993777791659037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,2560,0.0173342227935791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,64,0.0038124442928367188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,128,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,128,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,32,0.003903111235962974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,128,0.009334222310119206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,64,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,2048,0.02198666665289137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,32,0.007872888611422645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,65536,0.2843635612063938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,51200,0.2163848876953125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,65536,0.22394222683376738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,65536,0.1257902251349555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,51200,0.18270933628082275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,16384,0.07570577992333306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,256,0.006649777707126405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,51200,0.10098844766616821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,16384,0.05980889002482096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,12288,0.058740443653530545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,16384,0.08562311198976304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,12288,0.04509600003560384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,10240,0.04663377669122484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,12288,0.0299288895395067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,12288,0.06839199860890706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,65536,0.31059466467963326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,8192,0.039684444665908813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,10240,0.025898666845427618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,51200,0.2450408935546875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,10240,0.03874311182234023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,7168,0.03532622257868449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,8192,0.02200711104604933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,8192,0.03353155652681986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,10240,0.058872891796959766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,6144,0.029618667231665716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,7168,0.028931554820802476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,16384,0.03783822059631348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,256,0.011707555916574268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,6144,0.02571733295917511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,6144,0.017792888813548617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,7168,0.04532444477081299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,4096,0.021946666969193354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,5120,0.016135111451148987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,5120,0.022005332840813532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,6144,0.041053334871927895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,3584,0.018904889623324077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,4096,0.0188737776544359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,4096,0.01441422270403968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,5120,0.036008887820773654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,3072,0.016499555773205228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,3584,0.016861332787407767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,4096,0.03068444463941786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,3072,0.015483554866578845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,3072,0.012504888905419244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,2560,0.013770666387346057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,3584,0.028463999430338543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,7168,0.01997777819633484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,2048,0.011897777517636618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,2560,0.011172444456153445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,8192,0.050471110476387866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,2560,0.013863110707865821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,3072,0.02623288830121358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,5120,0.026268444127506677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,1536,0.009763555394278632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,2048,0.010020444790522257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,2560,0.02383822202682495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,1024,0.008135110967689091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,2048,0.012496000362767113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,1536,0.008364444805516137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,768,0.006717333363162146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,1536,0.0107351111041175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,1024,0.008005333443482717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,2048,0.021319111188252766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,512,0.005486222190989389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,1024,0.008957333034939235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,768,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,768,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,256,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,1024,0.0155102229780621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,3584,0.013442666994200813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,512,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,768,0.013752889302041797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,512,0.007164444360468123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,128,0.0031591111587153543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,256,0.006647999915811751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,256,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,512,0.013784888717863293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,64,0.003022222262289789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,128,0.006694222076071634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,32,0.0031724443866146933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,128,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,256,0.011767999993430244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,64,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,128,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,32,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,65536,0.26307911343044704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,65536,0.20719644758436415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,51200,0.20629599359300402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,65536,0.12298311127556695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,1536,0.01886399918132358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,51200,0.16273421711391872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,51200,0.0985635585255093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,16384,0.036575112077924944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,16384,0.056109335687425405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,12288,0.05617955658170912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,16384,0.08581333027945624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,12288,0.028592887851927016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,10240,0.04663377669122484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,51200,0.24047199885050455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,65536,0.30704087681240505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,12288,0.06738311052322388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,8192,0.038013332419925265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,10240,0.03733155462476943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,10240,0.024681778417693243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,7168,0.033640000555250384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,8192,0.021261332763565913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,8192,0.03158400124973721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,10240,0.0581351121266683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,16384,0.072653333346049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,6144,0.02933599882655674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,7168,0.01924533314175076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,8192,0.04924355612860786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,7168,0.02754488918516371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,5120,0.02498133314980401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,6144,0.023789332972632513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,6144,0.017469333277808297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,4096,0.02088000045882331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,7168,0.044871111710866295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,5120,0.0158906661801868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,5120,0.02121066715982225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,12288,0.043099555704328746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,3584,0.018231110440360177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,4096,0.014143110977278816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,4096,0.017991110682487488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,5120,0.03512710995144314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,3072,0.016176889340082806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,3584,0.01310222182008955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,2560,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,3584,0.016407110624843173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,4096,0.03029600116941664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,3072,0.012067555553383298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,2048,0.011882666912343768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,3072,0.01499466598033905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,3584,0.02822222312291463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,2560,0.01072533345884747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,2560,0.01368977791733212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,1536,0.009619555539555019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,3072,0.025398222936524287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,2048,0.008984888593355814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,2048,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,2560,0.022671999202834234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,1024,0.0079297779334916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,1536,0.008011555506123437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,1536,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,768,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,1024,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,1024,0.00832266691658232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,1536,0.01793866687350803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,512,0.004944889081848992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,768,0.007160888777838812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,6144,0.04027377896838718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,768,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,256,0.004015999949640698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,1024,0.014708444476127625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,512,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,512,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,128,0.0035031110876136353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,256,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,768,0.013782222237851886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,64,0.0029582221888833572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,256,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,512,0.012787555654843649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,32,0.0032133333798911837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,256,0.01148088855875863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,65536,0.19349066416422525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,64,0.006298666612969504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,128,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,32,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,51200,0.14896978272332087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,65536,0.16565867265065512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,2048,0.021334222621387903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,65536,0.0764897796842787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,51200,0.1298124392827352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,16384,0.05238755544026693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,51200,0.06157955858442518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,16384,0.023910222782029047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,16384,0.04530488782458835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,12288,0.0410426656405131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,16384,0.05197777681880527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,12288,0.035043554173575506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,65536,0.17309955755869547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,51200,0.13804354932573107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,10240,0.033960888783137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,12288,0.01941688855489095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,12288,0.041783110962973706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,8192,0.02827111217710707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,10240,0.01699466672208574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,10240,0.030340443054835003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,7168,0.02482044365670946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,8192,0.01497333414024777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,10240,0.03706399930848016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,8192,0.032109333409203425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,7168,0.013806222213639153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,6144,0.02159822152720557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,7168,0.028488000233968098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,6144,0.019807111885812547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,5120,0.018195556269751657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,128,0.006648889018429651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,6144,0.01275288893116845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,128,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,6144,0.026766222384240892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,4096,0.015260444747077094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,5120,0.011676444775528379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,5120,0.01719911065366533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,3584,0.013451555536852943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,4096,0.0107004443804423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,4096,0.015211555692884656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,5120,0.02385955552260081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,3072,0.012048000262843238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,8192,0.02558044426971012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,4096,0.02167911165290409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,3584,0.009732444253232744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,3584,0.0140684445699056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,2560,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,7168,0.022389334109094407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,3072,0.009031111167536842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,3072,0.013083555632167392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,2048,0.009205333060688442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,3584,0.019711110326978896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,2560,0.008087111016114553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,2560,0.011907555162906647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,1536,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,3072,0.018896889355447557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,2048,0.0075039996041191956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,2048,0.010617777705192566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,2560,0.01756000022093455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,1024,0.005982222242487802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,1536,0.0070977773931291364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,768,0.004950222041871813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,2048,0.016170667277442086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,1024,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,512,0.003690666622585721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,1024,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,768,0.0063697778516345555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,768,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,1024,0.012069332930776807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,256,0.003165333428316646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,512,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,512,0.0064071110553211635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,768,0.011906666888131036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,128,0.00282222218811512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,256,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,512,0.010715555813577441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,256,0.006120000034570694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,64,0.0028035555862718155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,32,0.0031315556002987754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,128,0.006149333384301927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,128,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,256,0.010355555348926121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,64,0.006112888869312074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,65536,0.17878311210208467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,32,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,65536,0.15871466530693903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,51200,0.13917333549923366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,1536,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,65536,0.07249066564771864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,16384,0.048433777358796865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,1536,0.014478221535682678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,51200,0.125763561990526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,12288,0.03752977649370829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,51200,0.058509336577521436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,16384,0.022441777918073866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,16384,0.044461333089404635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,10240,0.031009776724709406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,65536,0.1649893389807807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,12288,0.018557333283954196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,12288,0.034028444025251604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,16384,0.051015999582078725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,8192,0.026491555902693007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,51200,0.12888888518015543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,10240,0.01646844380431705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,12288,0.04127111037572225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,7168,0.023223999473783705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,8192,0.014459555347760519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,128,0.00815288888083564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,10240,0.03591466612286038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,6144,0.01993777851263682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,8192,0.030936890178256567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,7168,0.021653332644038733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,7168,0.01350400017367469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,5120,0.017122666041056316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,7168,0.028456889920764502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,6144,0.012395555774370829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,6144,0.019320888651741874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,5120,0.01719199948840671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,5120,0.011415111521879831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,6144,0.02553422252337138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,5120,0.023192000057962205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,3584,0.012726222475369772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,4096,0.02114222281508976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,10240,0.029383109675513372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,3584,0.01386222243309021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,8192,0.024775111012988623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,3072,0.011516444385051727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,3584,0.009183111290136972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,3072,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,2560,0.01000355515215132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,3072,0.01832799944612715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,4096,0.014781332678265043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,2560,0.011759111450778114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,2048,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,2560,0.007732444339328342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,4096,0.00999022192425198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,4096,0.01497511068979899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,2560,0.01702666613790724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,1536,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,2048,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,1536,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,2048,0.015792888071801927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,1024,0.005700444595681296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,1536,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,1536,0.014125333891974555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,768,0.0047271110945277745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,1024,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,1024,0.01203733351495531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,512,0.0036799998746977914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,768,0.01143022212717268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,512,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,2048,0.00979911122057173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,256,0.0031902222997612427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,512,0.006169777777459886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,512,0.011369778050316704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,256,0.005975111077229182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,128,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,256,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,1024,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,3072,0.012824888858530255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,256,0.009775111244784461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,64,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,128,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,32,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,128,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,768,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,64,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,65536,0.1729351149664985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,128,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,32,0.006441777778996362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,51200,0.13580177889929876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,65536,0.12254399723476833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,65536,0.06997244225607978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,16384,0.04593955477078756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,51200,0.09478844536675347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,51200,0.05670844184027778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,12288,0.03571910990609063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,16384,0.03501688771777683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,16384,0.02200711104604933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,51200,0.12097511026594375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,65536,0.15037866433461508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,10240,0.029418667157491047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,16384,0.04856355653868782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,12288,0.026791999737421673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,3584,0.01973600023322635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,8192,0.024273778001467388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,10240,0.01588888963063558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,12288,0.038554665115144514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,10240,0.02293511066171858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,768,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,8192,0.01978844404220581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,8192,0.014168888330459595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,6144,0.018516444497638278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,10240,0.034210665358437434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,7168,0.01308444473478529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,8192,0.029901332325405542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,5120,0.015963556038008798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,7168,0.01749955614407857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,6144,0.012030222349696688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,6144,0.016193777322769165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,7168,0.026576888230111863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,4096,0.013208000196350945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,5120,0.010999110837777456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,5120,0.01421866648726993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,6144,0.02453688946035173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,3584,0.011622221933470832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,4096,0.009602666729026372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,4096,0.012675555215941535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,5120,0.022392888863881428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,3072,0.010029333333174387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,3584,0.011657777759763928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,3584,0.008928889201747047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,3584,0.018752889500723947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,3072,0.008240888516108194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,2560,0.008958222137557136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,7168,0.021372444099850122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,3072,0.010987555815113915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,2048,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,2560,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,3072,0.017831999394628737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,2560,0.009758222434255812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,1536,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,2048,0.008620444271299574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,2560,0.016466667254765827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,1536,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,1024,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,1536,0.01276622215906779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,12288,0.01791466606987847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,1024,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,4096,0.020121778051058452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,768,0.004179555508825514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,1024,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,768,0.006312888943486744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,768,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,512,0.0033724444607893624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,1024,0.01146666705608368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,768,0.010967999696731567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,256,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,512,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,512,0.006115555349323485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,2048,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,128,0.0025235555238193935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,512,0.010749333434634738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,256,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,1536,0.006999111009968652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,64,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,128,0.005886222339338726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,256,0.009640889035330879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,32,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,64,0.005629333357016246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,65536,0.133516444100274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,128,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,32,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,51200,0.11221244600084092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,65536,0.06762133042017619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,65536,0.12036000357733832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,16384,0.046318223079045616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,51200,0.09189866648779975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,51200,0.0540924436516232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,16384,0.03408177693684896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,65536,0.14903199672698975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,256,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,16384,0.020946666598320007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,12288,0.03557422094874912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,51200,0.11951111422644721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,2048,0.01443199978934394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,10240,0.02976800004641215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,128,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,12288,0.01719555589887831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,8192,0.024877332978778418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,16384,0.046967999802695386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,10240,0.015610666738616096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,12288,0.02578488820128971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,7168,0.0217902229891883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,10240,0.022516444325447083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,8192,0.013816888961527081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,12288,0.03797600004408095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,8192,0.0192248887485928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,7168,0.012730666332774691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,10240,0.03268266717592875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,5120,0.015934222274356417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,7168,0.017197334104114108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,8192,0.028600888119803533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,6144,0.011701333026091257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,4096,0.013359111216333179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,6144,0.015559110376569959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,7168,0.026391999589072332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,5120,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,5120,0.01384800010257297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,3584,0.01184622198343277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,4096,0.008974221845467886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,4096,0.012440000143316058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,5120,0.021900445222854614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,3072,0.010682666467295753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,3584,0.008635555704434713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,3584,0.011367999845080905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,4096,0.019645333290100098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,2560,0.009345778160625035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,3072,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,3072,0.010828444527255164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,2048,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,2560,0.00794044468137953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,2560,0.009705777797434065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,3072,0.017487999465730455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,1536,0.006559111177921295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,2560,0.016132444143295288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,2048,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,6144,0.01850222216712104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,1536,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,2048,0.013821333646774292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,1536,0.0123857781291008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,6144,0.024189333120981853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,1024,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,768,0.0039013334446483185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,1024,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,3584,0.018585777944988675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,1024,0.011372444530328115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,512,0.003504000190231535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,768,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,768,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,2048,0.00813333359029558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,512,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,512,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,256,0.0028524444335036804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,1536,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,1024,0.005010666532648934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,128,0.0025351111673646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,512,0.010038221875826517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,256,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,64,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,128,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,256,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,32,0.0024622221373849446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,64,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,128,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,32,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,768,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,51200,0.09211022324032253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,65536,0.06492088900672065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,65536,0.11130221684773763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,16384,0.03433510992262099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,51200,0.08762666914198135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,65536,0.14794577492607966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,256,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,16384,0.03137866655985514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,51200,0.1186151107152303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,16384,0.020247111717859905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,12288,0.029836446046829224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,10240,0.025096888343493145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,12288,0.016929777132140267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,128,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,16384,0.046033776468700834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,12288,0.024658666716681585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,8192,0.020403555697864957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,10240,0.02146133283774058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,10240,0.01510933372709486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,12288,0.036883556180530123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,7168,0.01792977750301361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,8192,0.013442666994200813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,8192,0.018668444620238412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,10240,0.03239555491341485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,6144,0.015404444601800708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,7168,0.012223111258612739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,7168,0.016928889685206942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,8192,0.027962668074501887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,5120,0.013321778012646569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,6144,0.011009777585665384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,51200,0.05205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,7168,0.025959110922283594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,4096,0.011594666375054253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,5120,0.013468444347381592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,6144,0.023803555303149756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,4096,0.011798222031858234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,4096,0.008643555144468943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,3584,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,4096,0.01925155520439148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,3584,0.008271999657154083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,3072,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,65536,0.10672711001502143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,3072,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,3584,0.01774133410718706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,2560,0.006911110960774952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,3072,0.016359110673268635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,2560,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,2048,0.006108444597986009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,2560,0.0075902218619982404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,6144,0.01516533394654592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,2560,0.014785778191354541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,5120,0.009597332941161262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,1536,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,2048,0.007243555453088548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,2048,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,1024,0.0038328886859946777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,5120,0.021515554851955835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,1536,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,1536,0.007000889215204451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,2048,0.014088888963063559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,1024,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,1536,0.012418666647540199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,1024,0.00656355544924736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,3584,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,512,0.0028897778441508612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,768,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,768,0.006095110956165526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,1024,0.011729777687125735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,256,0.0025457777082920074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,3072,0.009482666850090027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,512,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,512,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,768,0.010457778142558204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,128,0.002448888909485605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,256,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,512,0.010024000373151567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,64,0.0025031111306614345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,128,0.005716444303592046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,128,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,256,0.009915555516878763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,128,0.007347555624114142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,64,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,32,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,768,0.0035617777870761026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,51200,0.08967910872565375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,65536,0.10824000173144871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,65536,0.1476071145799425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,51200,0.08522755569881862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,16384,0.03991377684805129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,256,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,51200,0.05087644524044461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,51200,0.11786311202579075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,32,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,12288,0.030437333716286555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,16384,0.01995022263791826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,16384,0.03142133355140686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,65536,0.1021066639158461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,10240,0.022794667217466567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,12288,0.016535111599498324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,16384,0.045774221420288086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,12288,0.024111110303137038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,8192,0.017978666557206046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,10240,0.014880888991885714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,10240,0.021214221914609272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,12288,0.036728888750076294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,7168,0.016015110744370353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,65536,0.06390577554702759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,8192,0.013014222184816996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,10240,0.032115555471844144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,8192,0.018233777748213876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,6144,0.014278221461508008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,7168,0.01163200040658315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,5120,0.012434666355450949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,7168,0.016504888733228047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,8192,0.028070221344629925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,6144,0.010452444354693094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,4096,0.010844444235165914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,6144,0.014860444598727755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,7168,0.02574933403068119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,5120,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,3584,0.009959110783206092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,5120,0.01313066648112403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,6144,0.023425777753194172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,4096,0.008636444807052612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,3072,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,4096,0.01130311108297772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,5120,0.021288888321982488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,2560,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,3584,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,3584,0.010419555836253695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,4096,0.019131556153297424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,2048,0.005724444571468566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,3072,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,3584,0.017887110511461895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,2560,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,1536,0.005020444591840108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,2560,0.008735111190213097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,3072,0.01591466698381636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,2048,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,2048,0.008033778104517195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,1024,0.0041893331540955436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,1536,0.006785777707894643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,2560,0.014615111880832247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,1536,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,768,0.003342222215400802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,2048,0.014095111025704278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,1024,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,1024,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,1536,0.012192000117566852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,512,0.002955555501911375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,768,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,256,0.0027555556346972785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,768,0.006191111273235745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,512,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,512,0.00565155554148886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,768,0.010780444575680627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,128,0.002528888897763358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,256,0.005708444449636672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,256,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,512,0.010048000348938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,64,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,32,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,3072,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,128,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,256,0.009850666754775578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,128,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,65536,0.10662577549616496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,64,0.005535111245181825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,128,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,32,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,51200,0.08373955885569255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,65536,0.07000088691711426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,65536,0.06283733579847547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,16384,0.027829332484139338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,51200,0.04993066522810194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,51200,0.05508888761202494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,12288,0.01983022193113963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,65536,0.14790311124589708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,16384,0.019727999965349834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,16384,0.020959999826219346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,10240,0.016911110944218107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,1024,0.011055111057228513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,51200,0.11756977770063613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,12288,0.016428444120619033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,12288,0.016910221841600206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,16384,0.045258667733934194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,8192,0.014197332991494073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,10240,0.014565333724021912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,7168,0.012768000364303589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,10240,0.01513777838812934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,8192,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,12288,0.03625511129697164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,6144,0.011307555768224927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,8192,0.013450666434235044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,10240,0.03170933326085409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,7168,0.011032000184059143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,5120,0.009838222629494136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,7168,0.012364444633324942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,8192,0.027701333165168762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,6144,0.010400888820489248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,6144,0.011154666543006897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,4096,0.008731555607583787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,5120,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,7168,0.02544622288809882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,3584,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,6144,0.023212444451120164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,4096,0.00869066682126787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,4096,0.00869955536392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,5120,0.02113688819938236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,3072,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,3584,0.0083137775460879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,4096,0.018497778309716117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,3584,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,2560,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,3072,0.007858666280905405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,3584,0.01776266594727834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,2048,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,3072,0.016185777054892648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,2560,0.007139555282062954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,2560,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,1536,0.004569777597983678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,2560,0.014294221997261047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,2048,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,2048,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,1024,0.0034888887570963967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,1536,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,1536,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,2048,0.013784888717863293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,768,0.0031928889867332247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,1024,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,5120,0.0099982221921285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,1536,0.012279111478063794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,512,0.0029813332690132987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,768,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,768,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,1024,0.011086222198274402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,256,0.0025066667132907445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,512,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,512,0.005341333233647876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,128,0.0021928888228204516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,3072,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,512,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,256,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,64,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,128,0.0058311112225055695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,128,0.005136000199450387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,32,0.0021484444538752237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,65536,0.07481155792872111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,128,0.007672000262472365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,32,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,1024,0.0063271112740039825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,65536,0.067193779680464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,65536,0.06135644515355428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,51200,0.05935111310746935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,65536,0.14722843964894614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,51200,0.052898665269215904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,51200,0.04849244488610161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,16384,0.024104000793562994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,768,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,12288,0.022256000174416438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,256,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,16384,0.01959377858373854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,16384,0.020230222079488967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,51200,0.11748266220092773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,256,0.009673777553770278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,12288,0.015885333220163982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,12288,0.016522667474216886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,16384,0.04475733306672838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,10240,0.014022222823566861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,10240,0.014950222439236112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,12288,0.03586311141649882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,8192,0.011744000017642975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,10240,0.031758222315046526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,8192,0.013089777694808112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,8192,0.02755466600259145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,7168,0.012075555821259817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,6144,0.010801778071456485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,7168,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,7168,0.025400888588693406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,6144,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,6144,0.010032888915803697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,5120,0.009538667069541084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,10240,0.01960000064637926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,6144,0.02306488818592495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,64,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,4096,0.008552889029184977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,5120,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,5120,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,3584,0.007695111135641734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,4096,0.008615111311276754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,5120,0.02088800072669983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,7168,0.01219466659757826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,3072,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,4096,0.01800622211562263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,3584,0.00831911133395301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,3584,0.00794044468137953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,2560,0.006213333457708359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,3584,0.016756445169448853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,3072,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,3072,0.007344889144102733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,2048,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,2560,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,3072,0.015253333581818474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,1536,0.004237333519591225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,2048,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,2048,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,2560,0.014352889524565803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,1024,0.0034977777136696708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,1536,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,2048,0.013225778109497495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,1536,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,768,0.0031991110493739447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,1024,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,1024,0.0058151111006736755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,1536,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,4096,0.008309333688682979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,768,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,768,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,1024,0.011392888923486074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,256,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,512,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,512,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,768,0.01071111112833023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,128,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,256,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,256,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,512,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,64,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,128,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,128,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,32,0.002398222270939085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,256,0.009308444129096137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,65536,0.05947999821768867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,8192,0.013889777991506787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,128,0.007699555820888943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,32,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,51200,0.05023555623160469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,65536,0.06446311208936903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,65536,0.05973955657747057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,512,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,16384,0.020523554748959012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,51200,0.047275556458367236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,51200,0.050626665353775024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,65536,0.14708267317877874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,12288,0.017709333035680983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,16384,0.01926666663752662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,16384,0.020012444920010038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,51200,0.11735288302103679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,10240,0.015052444405025907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,12288,0.01516355574131012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,12288,0.016173332929611206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,16384,0.04434400134616428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,2560,0.007000000112586551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,10240,0.013064000341627332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,12288,0.035396443472968206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,10240,0.014676445060306124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,7168,0.012162666353914471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,8192,0.012708444562223224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,8192,0.011757333245542316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,10240,0.03128266665670607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,6144,0.010943111446168689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,64,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,7168,0.010713777608341642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,7168,0.011373333632946014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,8192,0.027455111344655354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,5120,0.00961511085430781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,6144,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,4096,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,7168,0.02493688960870107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,6144,0.010064000056849586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,5120,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,5120,0.008988444175985124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,6144,0.021985777550273474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,4096,0.008649778034951951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,4096,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,5120,0.020670221911536325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,3584,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,4096,0.017645332548353408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,3584,0.007892444729804993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,8192,0.01367200000418557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,2560,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,3584,0.016427555018001132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,3072,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,3072,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,2048,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,2560,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,3072,0.015495111544926962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,2560,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,1536,0.0042204442951414315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,2048,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,2048,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,2560,0.014333332578341166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,1024,0.0035111109415690103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,1536,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,2048,0.013120888835854001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,3584,0.007544000115659501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,1024,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,768,0.0031582222630580268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,1536,0.012096888489193387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,512,0.002867555452717675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,768,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,3072,0.006599111275540457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,1024,0.0107351111041175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,768,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,256,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,512,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,512,0.005283555636803309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,768,0.011017777853541903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,128,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,256,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,256,0.004956444518433677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,512,0.009639999932712978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,64,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,32,0.0021475555582178966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,128,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,128,0.005231110999981563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,256,0.009451555709044138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,65536,0.04853689008288913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,128,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,64,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,32,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,51200,0.04028977619277106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,1536,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,65536,0.06174666351742215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,16384,0.01554666625128852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,51200,0.04863022102249993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,65536,0.14668088489108616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,16384,0.019329778022236295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,16384,0.01856355534659492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,51200,0.11670755015479194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,16384,0.04422310988108317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,10240,0.012667555775907306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,12288,0.014135110709402295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,12288,0.015779554843902588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,8192,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,10240,0.012755555411179861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,12288,0.03513333201408386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,8192,0.011796444654464722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,10240,0.030974222554100886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,7168,0.011677333050303988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,8192,0.027104889353116352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,65536,0.05809599823421902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,6144,0.010414222048388587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,7168,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,51200,0.0468444459968143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,7168,0.0247688889503479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,12288,0.014006222287813822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,5120,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,6144,0.010031999813185798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,6144,0.009857777920034196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,4096,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,5120,0.009375111096435124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,6144,0.022291556000709534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,4096,0.00810933361450831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,4096,0.008640000389681922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,3584,0.006909333169460297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,10240,0.014099554883109199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,4096,0.017533333765135873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,3584,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,8192,0.011348444554540845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,3072,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,3584,0.007719111111429002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,2560,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,3072,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,3584,0.016188444362746347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,3072,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,2048,0.004562666846646203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,7168,0.010682666467295753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,2560,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,2560,0.007481777833567725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,3072,0.01556799974706438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,1536,0.0038559999730851916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,2048,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,2048,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,2560,0.014131555954615274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,1024,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,5120,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,1536,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,1536,0.006719111154476802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,2048,0.013071999781661563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,768,0.0028568889117903183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,5120,0.01977511081430647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,1024,0.006550222221348021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,1024,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,1536,0.012063110868136087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,512,0.002551111082235972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,768,0.0063368889192740125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,256,0.0025066667132907445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,512,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,768,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,512,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,128,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,256,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,256,0.0057884446448749965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,512,0.010018666585286459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,64,0.0019271110908852683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,128,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,256,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,32,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,65536,0.03545422355333964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,64,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,128,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,32,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,51200,0.02956533432006836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,65536,0.05957688887914022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,65536,0.057251552740732826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,768,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,51200,0.04602311054865519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,51200,0.04674222071965536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,65536,0.14636266231536865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,1024,0.011066666907734342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,12288,0.011702222128709158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,16384,0.016859556237856548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,51200,0.11607110500335693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,16384,0.01885066595342424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,10240,0.010623999767833285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,12288,0.014817777607176038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,12288,0.014144000079896716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,16384,0.04414222306675381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,8192,0.009379555781682333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,10240,0.01311644415060679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,12288,0.0351893338892195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,7168,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,8192,0.011413333316644033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,10240,0.029839999145931665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,8192,0.02604888876279195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,7168,0.010636444720957013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,6144,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,16384,0.013413333230548434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,7168,0.023960888385772705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,5120,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,6144,0.009639110830095079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,6144,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,6144,0.02166844407717387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,4096,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,5120,0.008882666627566019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,128,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,5120,0.019479110836982727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,10240,0.012676444318559436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,3584,0.006552888701359431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,4096,0.008039111064540016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,8192,0.0114720000161065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,3072,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,3584,0.007676444119877285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,7168,0.010562666588359408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,3584,0.016180444094869826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,2560,0.005197333378924264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,3072,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,3072,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,3072,0.01518311103185018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,2560,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,2048,0.004536889079544279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,5120,0.009315555294354757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,2560,0.014456889695591398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,1536,0.0038986665507157645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,2048,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,2048,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,4096,0.008384000096056197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,2048,0.013134222063753339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,1536,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,1536,0.006065777606434292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,3584,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,768,0.0029768887907266617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,1024,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,1024,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,768,0.00599911105301645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,512,0.0026968889352348116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,768,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,2560,0.007287110719415877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,768,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,512,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,512,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,256,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,512,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,256,0.005299555758635203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,1024,0.00319377767542998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,256,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,256,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,4096,0.01753422286775377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,64,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,1536,0.01202933324707879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,128,0.00563822231358952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,32,0.0021111110432280437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,65536,0.028824888997607764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,128,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,1024,0.011011555790901184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,64,0.004963555683692296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,51200,0.022745778163274128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,65536,0.05686577823426989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,65536,0.05786133474773831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,51200,0.04591111010975308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,51200,0.04589066571659512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,16384,0.011640888949235281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,65536,0.1459653377532959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,128,0.002161777681774563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,51200,0.11613244480556911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,16384,0.016703999704784818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,12288,0.010004444254769219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,16384,0.04387733340263367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,12288,0.013870221873124441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,128,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,12288,0.014513777361975776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,10240,0.012746666868527731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,12288,0.03428355521625943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,10240,0.01291822228166792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,8192,0.010321777727868822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,32,0.004965333475006951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,7168,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,8192,0.011284444067213269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,10240,0.03070488903257582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,8192,0.01142133358452055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,6144,0.008672888908121321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,7168,0.010399111443095738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,7168,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,8192,0.02600355611907111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,5120,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,6144,0.009741333623727163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,7168,0.02380888826317257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,6144,0.009661333428488838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,16384,0.018221333622932434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,5120,0.008902221918106079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,6144,0.021362667282422382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,3584,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,10240,0.01145244472556644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,4096,0.00829777783817715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,4096,0.008012444608741337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,5120,0.01960711181163788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,3072,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,4096,0.017151110702090792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,3584,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,2560,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,3072,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,3072,0.0076248885856734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,3584,0.01629866659641266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,2048,0.004578666554556953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,3072,0.015146666102939181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,2560,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,2048,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,2048,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,1536,0.0038951109680864546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,2560,0.014451555079884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,4096,0.0069057775868309875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,1024,0.003148444410827425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,1536,0.006724444528420766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,1536,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,1024,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,1536,0.012073777616024017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,768,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,1024,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,3584,0.007948444121413762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,512,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,768,0.006042666733264923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,768,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,512,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,768,0.010411555568377176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,512,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,256,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,512,0.010313777459992303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,128,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,256,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,256,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,2048,0.013069333301650153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,64,0.001793777777089013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,128,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,768,32,0.001814222170246972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,128,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,65536,0.023208000593715247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,64,0.005006222261322869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,1024,0.011402666568756104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,32,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,65536,0.056518223550584584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,65536,0.05673244264390734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,51200,0.01891644464598762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,256,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,65536,0.145632889535692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,16384,0.01216533366176817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,51200,0.045457777049806386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,51200,0.04455555478731791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,128,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,16384,0.016673778494199116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,16384,0.04323644439379374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,12288,0.014367111855083041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,12288,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,10240,0.00963199966483646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,2560,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,12288,0.03427111109097799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,10240,0.012776888906955719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,8192,0.008187555604510838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,10240,0.029483556747436523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,5120,0.00905688852071762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,8192,0.01131377783086565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,12288,0.010437332921557956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,7168,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,8192,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,7168,0.010374222364690568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,6144,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,8192,0.025683555338117812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,7168,0.023687111006842718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,6144,0.00959288908375634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,5120,0.006771555377377405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,6144,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,6144,0.020936000678274367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,5120,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,4096,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,5120,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,5120,0.01921333372592926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,51200,0.11537688308291966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,4096,0.008378666308191087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,3584,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,16384,0.017447999782032438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,3072,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,4096,0.017031111650996737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,3584,0.007962666451931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,3584,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,2560,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,7168,0.010715555813577441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,3072,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,3072,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,2048,0.004504000147183736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,3584,0.016224000189039443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,2560,0.007289778027269576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,2560,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,1536,0.0038782221575578055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,2048,0.007053333024183909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,2048,0.0063893331421746155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,2560,0.01442577772670322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,1024,0.003162666741344664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,10240,0.012841777669058906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,2048,0.013108444710572561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,1536,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,1536,0.006607111129495833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,768,0.0028533333291610083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,1024,0.006516444600290722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,1024,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,4096,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,512,0.002463999928699599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,768,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,768,0.005547555370463266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,1024,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,256,0.002190222135848469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,512,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,768,0.01038933296998342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,512,0.0053022222386466126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,128,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,256,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,256,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,512,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,128,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,128,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,3072,0.014768888552983602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,256,0.009376889301670922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,32,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,65536,0.019425777925385367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,128,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,64,0.005063999858167437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,51200,0.016555555992656283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,65536,0.051853332254621715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,65536,0.05585600270165337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,16384,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,51200,0.0398871103922526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,51200,0.04488799969355265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,1536,0.011715555356608497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,65536,0.14519111315409342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,12288,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,16384,0.016619554824299283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,51200,0.11523821618821885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,16384,0.01719199948840671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,10240,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,12288,0.014111111561457316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,12288,0.01423555612564087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,16384,0.04212355613708496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,8192,0.006016888966162999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,10240,0.012715555727481842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,512,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,10240,0.012796444197495779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,7168,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,10240,0.029513776302337646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,8192,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,8192,0.011022222538789114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,6144,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,7168,0.010416888528399997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,7168,0.01035377797153261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,32,0.005079999979999331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,8192,0.026001777913835313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,5120,0.006671111202902264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,6144,0.009764444496896531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,6144,0.0094373333785269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,7168,0.02346399923165639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,4096,0.006015111174848344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,5120,0.008640000389681922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,5120,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,3584,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,6144,0.020976000361972388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,5120,0.019311999281247456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,4096,0.008473777936564552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,4096,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,3072,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,4096,0.01681600014368693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,3584,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,3584,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,2560,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,12288,0.033598221010631986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,3072,0.007245333658324347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,3072,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,2048,0.004253333227501975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,2560,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,3072,0.015445333388116626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,2560,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,1536,0.0037368887828456033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,2048,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,2048,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,2560,0.014215999179416232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,1536,0.006650666809744305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,2048,0.012925333446926542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,1536,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,768,0.0028124445428450904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,1024,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,1536,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,768,0.005283555636803309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,768,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,1024,0.010683555569913654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,512,0.002467555511328909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,768,0.010399999717871347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,256,0.0021688888470331826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,512,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,3584,0.015838222371207345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,128,0.001998222122589747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,256,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,512,0.010041777458455827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,256,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,64,0.0018293333964215384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,1024,0.0031600000543726813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,128,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,256,0.009648000200589498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,65536,0.01292799992693795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,64,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,128,0.0075902218619982404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,1024,0.005766222046481238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,51200,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,65536,0.053730666637420654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,65536,0.0492142207092709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,51200,0.03965333435270522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,16384,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,65536,0.14491821659935847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,512,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,16384,0.016506666938463848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,51200,0.11530666881137425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,12288,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,16384,0.01723288827472263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,128,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,12288,0.014109333356221518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,256,32,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,12288,0.014161777165200977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,10240,0.012710221939616732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,12288,0.033612443341149226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,8192,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,32,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,10240,0.012740444805887012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,7168,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,8192,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,10240,0.029439111550649006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,8192,0.010757333702511258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,6144,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,51200,0.04297955499755012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,7168,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,7168,0.010399111443095738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,5120,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,8192,0.02574311031235589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,6144,0.009711110757456886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,6144,0.009039999710188972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,7168,0.023071999351183575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,4096,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,5120,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,5120,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,6144,0.021292444732454088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,3584,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,5120,0.01923822197649214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,10240,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,4096,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,3072,0.004971555537647671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,3584,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,3584,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,4096,0.01739911072784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,2560,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,3072,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,3584,0.01647555496957567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,3072,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,2048,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,2560,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,2560,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,3072,0.015101333459218344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,1536,0.003555555724435382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,2048,0.0064080001579390625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,2048,0.0070977773931291364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,2560,0.014144000079896716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,1024,0.002945777856641345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,1536,0.00665866666369968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,1536,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,2048,0.013210666676362356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,768,0.002805333377586471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,16384,0.04243555665016174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,4096,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,1024,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,1024,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,1536,0.012412444584899478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,768,0.005971555494599872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,768,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,256,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,512,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,768,0.010400888820489248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,512,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,128,0.0018835556175973679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,256,0.005003555367390315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,128,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,256,0.008974221845467886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,128,0.004959110998445087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,32,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,64,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,65536,0.009321778184837764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,32,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,512,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,16384,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,1024,0.011191999746693505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,65536,0.05266044537226359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,12288,0.005630222045713001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,512,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,10240,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,51200,0.04210755560133192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,8192,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,16384,0.016879111528396606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,7168,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,6144,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,12288,0.013847110999955071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,8192,0.010981332924630908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,7168,0.010048000348938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,128,64,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,4096,0.004584889031118817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,3584,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,5120,0.008407110969225565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,3072,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,4096,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,2560,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,3584,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,51200,0.010991999672518836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,2048,0.004810666872395409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,1536,0.003524444583389494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,3072,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,1024,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,2560,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,768,0.002855111120475663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,2048,0.006298666612969504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,1536,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,1024,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,256,0.0021253333737452826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,768,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,128,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,10240,0.012474666866991254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,512,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,64,0.0018177777528762817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,32,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,256,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,65536,0.013637333280510373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,6144,0.00944266633854972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,128,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,51200,0.011885333392355176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,64,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,16384,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,12288,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,32,0.004853333450025982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,65536,0.052503112289640635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,10240,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,51200,0.04244800077544319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,8192,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,12288,0.014144889182514615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,7168,0.005537777725193236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,10240,0.012551111479600271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,6144,0.004967111266321606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,8192,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,512,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,5120,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,4096,0.004563555535342959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,64,5120,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,5120,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,3584,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,3072,0.004598222259018156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,4096,0.007677333222495184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,2560,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,3584,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,2048,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,3072,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,1536,0.003519999898142285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,2560,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,1024,0.0028355556229750314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,16384,0.016884444488419425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,2048,0.006311999840868845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,768,0.0025519999778932994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,512,0.002537777854336633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,1536,0.005642666584915585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,256,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,1024,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,128,0.0018924443672100704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,768,0.005373333477311664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,64,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,7168,0.010336000058386061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,256,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,192,32,32,0.0018053334206342697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,128,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,6144,0.009330666727489894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,64,0.004966222163703707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,32,0.004850666556093427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,16384,0.386369784673055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,16384,0.27762044800652397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,12288,0.2912435531616211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,16384,0.13755288388994005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,51200,0.8417955504523383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,512,0.005113777601056629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,51200,1.2033910751342773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,10240,0.23534043629964194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,12288,0.20122667153676352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,51200,0.4077431095971002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,8192,0.0710560017161899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,12288,0.10383911265267266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,8192,0.13913867208692762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,8192,0.19263733757866752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,10240,0.1734373304578993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,7168,0.17776532967885336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,6144,0.14373421669006348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,5120,0.12950400511423746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,7168,0.1225022210015191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,6144,0.05558222532272339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,6144,0.10336711009343465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,4096,0.09970666964848836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,5120,0.08744355705049302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,4096,0.039777778916888766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,5120,0.051015999582078725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,4096,0.07583466503355238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,3072,0.07878044578764173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,3584,0.035261332988739014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,3584,0.06322844160927667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,2560,0.06357777780956693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,3072,0.03199288911289639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,3072,0.054994665914111666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,2048,0.051332443952560425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,10240,0.08650577730602688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,2560,0.02754577828778161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,2560,0.04853511187765333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,2048,0.024162666665183172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,2048,0.04071111149258084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,1024,0.028970665401882593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,7168,0.0630942251947191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,1536,0.018955555227067735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,1536,0.03234133455488417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,768,0.02181333303451538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,512,0.015531554818153381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,1024,0.015606222881211175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,1024,0.025247111916542053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,256,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,512,0.013268444273206921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,3584,0.08687822024027507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,768,0.024671110841963027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,512,0.018608889646000333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,128,0.008271999657154083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,64,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,256,0.011876444849703046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,128,0.011556444068749746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,32,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,128,0.014750222365061441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,64,0.01479466590616438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,32,0.015210666590266757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,1536,0.03997066617012024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,768,0.014084445105658637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,65536,0.8590879970126681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,65536,1.154334174262153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,256,0.015632889337009855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,51200,0.9067039489746094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,16384,0.29560356669955784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,51200,0.6768648889329699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,65536,0.40982309977213544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,16384,0.21525422732035318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,16384,0.1083128915892707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,12288,0.22274133894178602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,12288,0.08355289035373265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,10240,0.19965067174699572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,12288,0.16026666429307726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,51200,0.3249715434180366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,8192,0.1642577780617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,7168,0.14407732751634386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,10240,0.13878933588663736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,8192,0.05727377865049574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,6144,0.12354221608903672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,8192,0.11522932847340901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,7168,0.10121955474217732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,7168,0.05027911067008972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,6144,0.08362755510542129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,6144,0.044574220975240074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,4096,0.08510844575034247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,5120,0.07127733363045587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,3584,0.07567289140489367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,4096,0.03279022375742594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,4096,0.058008001910315625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,3072,0.0661768913269043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,3584,0.029162665208180744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,3584,0.051570667160881885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,2560,0.05581511391533745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,3072,0.026434666580624048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,3072,0.04529688755671183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,2048,0.04569955666859945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,2560,0.023593778411547344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,10240,0.06930489010281034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,2560,0.03916800022125244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,1536,0.0355751117070516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,1024,0.02583822276857164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,2048,0.019716444942686293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,2048,0.03346844514211019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,1536,0.016559999850061204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,768,0.019710222880045574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,1536,0.027095999982621934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,1024,0.01387644476360745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,5120,0.10471289025412665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,512,0.0144213338692983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,1024,0.020972443951500788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,256,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,768,0.019662222928471036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,768,0.012708444562223224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,5120,0.037985778517193265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,128,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,256,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,512,0.015475556254386902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,32,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,256,0.01328888866636488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,128,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,128,0.012773333324326409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,64,0.012456888953844706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,32,0.013184888495339287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,65536,0.3748186694251166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,51200,0.29447023073832196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,65536,0.30269156561957467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,16384,0.09613955683178371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,65536,0.14047111405266657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,51200,0.2349475489722358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,12288,0.07545955313576592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,51200,0.11119733916388617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,512,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,16384,0.04077777928776211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,10240,0.06284977992375691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,12288,0.03188000122706095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,16384,0.0793244441350301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,12288,0.059939556651645236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,8192,0.051402668158213295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,7168,0.04482044445143806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,10240,0.02747822304566701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,6144,0.03923199905289544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,8192,0.023601777023739282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,8192,0.04372088776694404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,7168,0.02109866672092014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,10240,0.05141689048873054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,5120,0.03312355611059401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,6144,0.018780445059140522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,4096,0.0270453327231937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,6144,0.033667554457982383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,5120,0.016961778203646343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,4096,0.01495822270711263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,4096,0.024490666058328416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,3072,0.02181688944498698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,3584,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,3584,0.022142221530278523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,2560,0.017978666557206046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,3072,0.012776888906955719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,64,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,3072,0.019360888335439894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,2048,0.014718222949239941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,2560,0.012067555553383298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,2560,0.016991110311614144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,1536,0.011726222104496427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,2048,0.010841777755154504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,2048,0.014860444598727755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,7168,0.037880887587865196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,1536,0.009012444151772393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,3584,0.02456088860829671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,1024,0.008064889245563084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,768,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,5120,0.028883556524912517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,1024,0.01070755554570092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,768,0.007479999628331926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,768,0.00998755544424057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,256,0.004200889004601372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,512,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,512,0.008027555214034187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,128,0.0037288889288902283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,256,0.006797333144479328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,256,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,64,0.003216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,128,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,128,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,32,0.0032213332338465583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,1024,0.009111111362775167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,64,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,1536,0.012727110750145383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,32,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,65536,0.2836719883812798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,51200,0.21693955527411568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,512,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,16384,0.07092444764243232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,65536,0.21951288647121855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,65536,0.1265653371810913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,51200,0.1719519959555732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,51200,0.10152800215615167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,12288,0.05577422512902153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,10240,0.04805066519313388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,16384,0.03752800159984165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,16384,0.057872891426086426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,12288,0.029706666866938274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,8192,0.0390035543176863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,7168,0.03485777642991807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,10240,0.025732444392310247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,10240,0.038093331787321306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,8192,0.021697777840826247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,6144,0.030003554291195337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,8192,0.032624890406926475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,7168,0.019859555694792006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,5120,0.025637333591779072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,7168,0.02870755394299825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,4096,0.022044445077578228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,6144,0.01789688898457421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,3584,0.018851555056042142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,5120,0.01605955594115787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,6144,0.025397333833906386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,5120,0.021726222501860723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,3072,0.016350222958458793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,3584,0.01331911070479287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,4096,0.018590221802393597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,2560,0.013753777576817406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,3584,0.01685422162214915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,3072,0.012472000386979846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,2048,0.01198755535814497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,3072,0.015248888068728976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,1536,0.009754666851626502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,2560,0.013759111364682516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,2048,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,1024,0.007902222375075022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,2048,0.012422222230169507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,1536,0.008358221915033128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,768,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,12288,0.04452622267935011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,1536,0.010689777632554373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,1024,0.008055110772450766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,512,0.005642666584915585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,1024,0.008620444271299574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,768,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,256,0.0041777777175108595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,512,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,768,0.00869688888390859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,128,0.003528888854715559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,512,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,64,0.003173333282272021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,256,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,128,0.006664000037643645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,32,0.003181333343187968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,4096,0.014448000325096978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,128,0.00664533343580034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,64,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,65536,0.2621777852376302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,2560,0.011535999675591787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,51200,0.2053626643286811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,65536,0.12241600619422065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,65536,0.20416000154283312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,16384,0.06875377893447876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,51200,0.0984613365597195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,51200,0.15991022851732042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,128,0.007377777662542131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,16384,0.05496444304784139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,10240,0.045884443653954395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,12288,0.028468443287743464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,12288,0.042388445801205106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,256,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,8192,0.03836444351408217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,10240,0.02455111179086897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,10240,0.036729776197009616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,7168,0.032765332195493914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,8192,0.021294222937689886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,8192,0.03108000093036228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,6144,0.029211554262373183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,7168,0.019341333044899833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,32,0.006986666884687211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,7168,0.027408889598316614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,5120,0.024832889437675476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,6144,0.017551110850440133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,4096,0.020649777518378366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,6144,0.023764444722069636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,5120,0.01556799974706438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,3584,0.017921777235137094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,4096,0.014124444789356656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,5120,0.020741333564122517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,3072,0.01612355477280087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,4096,0.017835555805100333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,12288,0.05291555656327141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,16384,0.036130666732788086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,2560,0.013754666679435305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,3584,0.013088888592190213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,2048,0.01180355581972334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,3584,0.016514667206340365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,3072,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,1536,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,3072,0.014914666612943014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,2048,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,1024,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,2560,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,2560,0.01328000012371275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,1536,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,2048,0.012071111136012606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,512,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,1024,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,768,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,1024,0.00813066628244188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,256,0.0041706665522522396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,768,0.007815111014578078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,512,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,128,0.0035253332720862497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,512,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,64,0.00350222239891688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,128,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,32,0.003549333247873518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,64,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,32,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,65536,0.21962577766842314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,768,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,1536,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,65536,0.1640577713648478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,65536,0.07557688819037543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,51200,0.17150133185916475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,256,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,51200,0.12714489301045737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,12288,0.045484443505605064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,256,0.006595555692911148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,16384,0.023706666297382776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,16384,0.0450817777050866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,128,0.0063786668082078295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,10240,0.03835555579927232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,8192,0.03147466646300422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,12288,0.019221333993805777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,7168,0.02758933272626665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,10240,0.016862221890025668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,12288,0.03466577662361993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,8192,0.014923555983437432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,10240,0.02974133359061347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,6144,0.023676445086797077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,8192,0.02529955572552151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,5120,0.020609777834680345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,7168,0.013767110804716745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,6144,0.01275022245115704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,7168,0.021998221675554912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,6144,0.019581332802772522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,3584,0.014872888724009195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,4096,0.010673777924643623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,4096,0.015055111712879606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,16384,0.057300441794925265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,3072,0.013306666579511432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,3584,0.009586667021115622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,3584,0.013778666655222574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,2560,0.011581333147154914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,51200,0.06058222055435181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,3072,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,2048,0.010025777750545079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,2560,0.008064000142945183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,3072,0.012876444392734103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,1536,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,2560,0.011648000114493899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,2048,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,1024,0.006765333314736684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,2048,0.01036977767944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,1536,0.008804444637563493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,768,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,1024,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,1024,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,512,0.004214222232500712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,4096,0.016975111431545682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,768,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,256,0.0032302221904198327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,5120,0.011805333197116852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,768,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,5120,0.01722133325205909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,128,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,512,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,64,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,256,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,32,0.0028248888750871024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,128,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,256,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,128,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,64,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,65536,0.20457511478000215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,32,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,51200,0.16156000561184353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,65536,0.07164266374376085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,65536,0.1570168866051568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,16384,0.05403022302521599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,1536,0.007134222322040134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,12288,0.04229066769282023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,51200,0.058493336041768394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,51200,0.12312089072333442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,10240,0.03548622131347656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,16384,0.04353511002328661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,12288,0.01847555571132236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,12288,0.03350933392842611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,8192,0.030293332205878362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,10240,0.016584888100624084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,7168,0.026191110412279766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,10240,0.02857422166400486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,8192,0.014511111709806653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,6144,0.022708444131745234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,512,0.006400888992680444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,7168,0.013430222041077085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,5120,0.019526221685939364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,8192,0.024309333827760484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,4096,0.016368000043763053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,7168,0.021351110604074266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,6144,0.012420444852775998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,3584,0.014208000567224292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,6144,0.019184889064894784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,4096,0.010247999595271217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,5120,0.01698844465944502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,3072,0.012835555606418185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,4096,0.01480888823668162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,3584,0.009076444639099969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,2560,0.01089777797460556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,3584,0.013794666363133324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,2048,0.009678222239017487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,3072,0.012753778033786349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,2560,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,1536,0.007880888879299164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,2560,0.011413333316644033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,16384,0.022668444448047217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,1024,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,2048,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,768,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,1536,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,1024,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,1536,0.008309333688682979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,512,0.0041991112132867174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,768,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,1024,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,256,0.0032142222755485107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,768,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,512,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,128,0.0028373334142896864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,256,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,512,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,64,0.003206222214632564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,256,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,128,0.005893333504597346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,32,0.0032222221295038858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,128,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,65536,0.16365688376956516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,64,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,32,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,3072,0.008273777862389883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,65536,0.11846221817864312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,51200,0.12250310844845241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,16384,0.04406933320893181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,65536,0.06961511241065131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,12288,0.03420711225933499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,51200,0.05622933308283488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,16384,0.021792888641357422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,2048,0.010058666268984476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,10240,0.030729777283138696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,16384,0.03406933281156752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,12288,0.018108444081412423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,12288,0.025983111725913152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,10240,0.01591999994383918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,7168,0.021891555852360193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,10240,0.02261688808600108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,8192,0.014135110709402295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,6144,0.01868355605337355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,8192,0.0192239996459749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,7168,0.01310222182008955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,7168,0.017103999853134155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,5120,0.011424889167149862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,5120,0.016233777006467182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,6144,0.012061333490742577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,4096,0.013743999931547375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,6144,0.015695111619101632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,3584,0.011974222130245633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,4096,0.009336888790130615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,5120,0.014039110806253223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,3072,0.010800888968838586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,4096,0.0124177775449223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,3584,0.008648888932334052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,2560,0.009413333402739631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,3584,0.011640888949235281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,3072,0.008162666526105668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,2048,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,3072,0.010666666759385003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,51200,0.09279377592934503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,2560,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,1536,0.0068897778789202375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,2560,0.009707555174827576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,8192,0.024915556112925213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,1024,0.005541333307822545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,2048,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,2048,0.008352888955010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,768,0.004367111043797599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,1536,0.0069208890199661255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,512,0.0035626664757728577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,1536,0.007342221836249034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,768,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,1024,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,256,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,512,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,768,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,128,0.002528888897763358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,512,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,256,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,64,0.0024480000138282776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,32,0.002475555572244856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,128,0.005879999862776862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,5120,0.010882666541470421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,128,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,64,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,65536,0.14487466547224256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,32,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,51200,0.11619199646843804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,65536,0.06695910957124498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,16384,0.042097777128219604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,51200,0.053992887338002525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,51200,0.09119377533594768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,12288,0.03314044409328037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,16384,0.020920000142521326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,1024,0.0064088888466358185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,12288,0.017313778400421143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,12288,0.025154666768179998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,8192,0.023415999280081853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,10240,0.015452444553375244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,10240,0.021911111142900255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,7168,0.020584000481499564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,8192,0.013627555635240344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,8192,0.018583999739752877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,6144,0.017603556315104168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,7168,0.012713777522246042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,5120,0.015219555960761176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,7168,0.016855110724767048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,6144,0.01143555591503779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,4096,0.012748444245921241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,65536,0.1166986624399821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,6144,0.015452444553375244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,5120,0.010050666828950247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,3584,0.011345778074529437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,5120,0.013753777576817406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,4096,0.008716444174448649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,3072,0.010330666270520952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,3584,0.008271111382378472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,4096,0.012136889000733694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,2560,0.008903111020723978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,16384,0.03284444411595663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,3072,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,10240,0.027633777923054163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,3072,0.010398222340477837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,2560,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,1536,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,2560,0.008817777865462834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,1024,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,2048,0.008059555457697975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,1536,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,768,0.004234666625658671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,256,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,1536,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,1024,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,512,0.0035342222286595237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,1024,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,768,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,256,0.0028808888875775864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,512,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,768,0.006312888943486744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,128,0.0025537777692079544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,512,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,256,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,64,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,256,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,128,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,32,0.0027928888383838865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,128,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,3584,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,65536,0.11743022335900201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,64,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,2048,0.007869333028793335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,32,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,51200,0.09258044428295559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,16384,0.03476266728507148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,65536,0.11025866534974839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,2048,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,65536,0.06446844339370728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,51200,0.05183910992410448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,12288,0.028074665202034846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,16384,0.020213334096802604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,10240,0.022815111610624526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,8192,0.018940443793932598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,16384,0.030983110268910725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,12288,0.016861332787407767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,7168,0.01629599928855896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,12288,0.0240000006225374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,10240,0.021135111649831135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,8192,0.013139555851618448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,6144,0.014367111855083041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,8192,0.018331555856598746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,7168,0.012048888537618848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,5120,0.012412444584899478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,7168,0.016398222910033334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,6144,0.011081777513027191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,4096,0.010879110958841113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,5120,0.010064000056849586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,6144,0.014848000473446317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,3584,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,4096,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,5120,0.013405332962671915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,3072,0.007851555943489075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,4096,0.011805333197116852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,3584,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,2560,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,2048,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,3584,0.010662222074137794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,3072,0.007920888562997183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,1536,0.004984000076850255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,2560,0.007548444800906711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,2048,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,2560,0.008778666456540426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,1024,0.004173333446184794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,2048,0.007939555578761632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,1536,0.0069066666894488865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,768,0.003565333369705412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,1536,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,512,0.0030453333424197305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,1024,0.006763555523422029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,768,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,1024,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,256,0.002552888873550627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,768,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,512,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,128,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,512,0.005660444498062134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,256,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,64,0.002510222295920054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,32,0.0025395556456512874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,128,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,128,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,65536,0.10179377926720513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,64,0.005640000104904175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,32,0.005630222045713001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,51200,0.08238933483759563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,65536,0.06334755818049113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,65536,0.1069013343916999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,16384,0.03271733389960395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,51200,0.050761779149373375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,51200,0.08370933267805312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,12288,0.02608888844648997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,51200,0.08545866939756606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,10240,0.021340444684028625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,16384,0.020064888728989493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,10240,0.015119999647140503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,16384,0.030264887544843886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,8192,0.017895110779338412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,12288,0.02377066678471035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,10240,0.014882667197121514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,7168,0.01588355501492818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,8192,0.012714666624863943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,10240,0.0207013338804245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,6144,0.014010666145218743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,8192,0.017715555098321702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,7168,0.011512888802422417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,256,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,5120,0.01220444424284829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,7168,0.01610311037964291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,6144,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,4096,0.010658666491508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,5120,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,6144,0.014815110299322339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,3584,0.009401777552233802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,5120,0.013094222380055321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,4096,0.008615111311276754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,3072,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,4096,0.011129777464601727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,3584,0.00832266691658232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,2560,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,3072,0.00981955561372969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,3584,0.010101333260536194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,2048,0.006542222367392645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,3072,0.009369778136412302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,1536,0.0053297777970631914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,2048,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,2560,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,2048,0.007740444607204861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,1024,0.004218666503826777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,12288,0.01664355562792884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,1536,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,1536,0.007227555745177799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,1024,0.006393777827421824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,512,0.003206222214632564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,1024,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,768,0.006235555642180973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,768,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,256,0.002806222273243798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,512,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,512,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,256,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,256,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,64,0.0022044444663657085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,128,0.005723555468850666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,32,0.002198222196764416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,128,0.005294222384691238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,3072,0.007701333198282454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,65536,0.08150844441519843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,2560,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,64,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,32,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,51200,0.06482577986187406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,16384,0.024846222665574815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,65536,0.062331557273864746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,768,0.00360355567600992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,51200,0.04965600040223864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,12288,0.020950223008791607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,51200,0.053527110152774386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,16384,0.019574221637513902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,10240,0.0177831103404363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,16384,0.02057066725360023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,12288,0.01640888883007897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,8192,0.015771556231710646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,10240,0.014485332700941296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,7168,0.014360000689824423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,128,0.002459555657373534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,8192,0.01310755560795466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,6144,0.01350488927629259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,7168,0.010718222293588849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,5120,0.011506666739781698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,7168,0.011761777930789523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,6144,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,4096,0.009311111436949836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,5120,0.009315555294354757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,6144,0.011062222222487131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,5120,0.009986666341622671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,4096,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,65536,0.06819199853473239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,3072,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,3584,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,3584,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,2560,0.006831999868154526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,3072,0.00796622203456031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,3072,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,12288,0.01684711045689053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,2048,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,2560,0.007605333295133378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,8192,0.012335110869672565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,10240,0.01482311056719886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,2560,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,1536,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,1024,0.003830222205983268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,2048,0.007165333463086023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,768,0.003375111147761345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,1536,0.00666933341158761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,1024,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,2048,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,512,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,1536,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,1024,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,256,0.0026142222599850762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,768,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,128,0.00250133333934678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,768,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,256,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,512,0.005520888914664586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,3584,0.00865955568022198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,64,0.0021964444054497615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,256,0.005207111024194294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,32,0.0023253332409593794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,128,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,65536,0.07426044676038954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,128,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,4096,0.008571555217107138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,51200,0.05918311410480075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,32,0.005450666778617435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,65536,0.06116000148985121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,16384,0.023758222659428913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,65536,0.06520889202753703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,51200,0.04855466551250882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,51200,0.05113777849409315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,16384,0.01937244501378801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,10240,0.01683822274208069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,16384,0.01993600030740102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,12288,0.015847111741701763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,8192,0.014175999495718213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,12288,0.01610933409796821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,10240,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,7168,0.012700444294346703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,8192,0.011396444506115384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,10240,0.01444533301724328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,512,0.005882666756709416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,6144,0.011029333704047732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,8192,0.012626666989591388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,7168,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,7168,0.011618666350841522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,6144,0.00998488896422916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,4096,0.00869422240389718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,6144,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,5120,0.009118222528033787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,64,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,5120,0.009026666482289633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,3584,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,3072,0.007247111035717859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,4096,0.008627555436558193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,4096,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,2560,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,3584,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,12288,0.019281778070661757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,3072,0.00795199970404307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,3584,0.007677333222495184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,2048,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,1536,0.004566222015354368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,2560,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,2560,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,1024,0.0038355555799272326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,2048,0.006581333362393909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,1536,0.006671999891599019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,768,0.003555555724435382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,1536,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,1024,0.006342222293217977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,512,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,1024,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,768,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,256,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,768,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,5120,0.009888889061080085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,128,0.0025315555847353404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,512,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,512,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,256,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,64,0.0021262222694026097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,128,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,32,0.0021848889688650766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,128,0.005092444519201915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,65536,0.05766489108403524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,64,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,51200,0.0469377769364251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,32,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,65536,0.05851733022265964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,3072,0.007400000260935889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,65536,0.06224977970123291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,12288,0.015840000576443143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,51200,0.04724444283379448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,2048,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,51200,0.04917066627078586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,16384,0.01905333333545261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,10240,0.015451555450757345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,8192,0.014215999179416232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,12288,0.015171556009186639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,16384,0.01940266622437371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,10240,0.013125333521101209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,12288,0.015800888339678448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,7168,0.014358222484588623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,10240,0.014008000493049622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,7168,0.010703110860453712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,8192,0.012140444583363004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,5120,0.009661333428488838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,7168,0.011056889262464313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,6144,0.010016000105275048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,6144,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,4096,0.008290666672918532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,256,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,5120,0.00908799966176351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,3584,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,5120,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,4096,0.008344000412358178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,3072,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,3584,0.008053333395057255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,4096,0.008162666526105668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,2560,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,16384,0.01979555520746443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,3072,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,3072,0.0074933336840735535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,2048,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,2560,0.007548444800906711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,1536,0.0041999999019834726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,2560,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,2048,0.006755555669466655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,1024,0.003479111111826367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,1536,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,6144,0.012696888711717395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,1536,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,1024,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,512,0.002801777794957161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,768,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,768,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,256,0.002471999989615546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,512,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,512,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,128,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,256,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,256,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,64,0.0021271111650599376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,128,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,32,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,128,0.005054222212897407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,2048,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,65536,0.045078221294615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,64,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,51200,0.03541510966089036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,32,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,65536,0.05796711312399971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,65536,0.05967999829186333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,51200,0.04645422101020813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,51200,0.047480000389946826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,12288,0.01290844463639789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,16384,0.0184568895234002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,16384,0.01891999940077464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,10240,0.011990221838156382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,12288,0.014133334159851074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,8192,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,12288,0.015453333655993143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,3584,0.007711110843552484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,7168,0.011558222273985544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,10240,0.01275466630856196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,10240,0.013630222115251752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,8192,0.011346666349305047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,6144,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,8192,0.011528000235557556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,7168,0.010680889089902243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,5120,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,7168,0.01039644413524204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,4096,0.007810666329330868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,6144,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,5120,0.009354666703277165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,8192,0.011351111034552256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,3584,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,768,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,5120,0.008811555802822113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,3072,0.0058355554938316345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,4096,0.008375999828179678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,4096,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,2560,0.005231110999981563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,3584,0.00814844419558843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,2048,0.004582222137186262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,3072,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,3584,0.007637333538797166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,2560,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,3072,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,2048,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,1024,0.0034426665968365143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,1536,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,1536,0.006235555642180973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,768,0.0028782222006056043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,1024,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,1024,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,512,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,768,0.006216889040337668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,768,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,6144,0.00980711148844825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,256,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,512,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,128,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,512,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,256,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,64,0.001855111163523462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,16384,0.016144000821643405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,128,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,32,0.0021102221475707158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,128,0.004948444250557157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,65536,0.033210668298933245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,64,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,1536,0.003945777813593547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,51200,0.026408889227443274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,65536,0.05719466341866387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,32,0.00509066672788726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,16384,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,2048,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,65536,0.05751466751098633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,12288,0.012103999654452005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,51200,0.04589688777923584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,51200,0.04534666736920675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,16384,0.01658933361371358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,10240,0.010948444406191507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,16384,0.018197332819302876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,12288,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,8192,0.01019377758105596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,12288,0.014109333356221518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,10240,0.012759110993809171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,7168,0.009553777674833933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,10240,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,6144,0.008719999757077958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,7168,0.010403555300500657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,8192,0.011032889286677042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,5120,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,7168,0.01015644437736935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,256,0.004986666556861666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,4096,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,6144,0.009463110731707679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,5120,0.008984888593355814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,3584,0.00630044440428416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,5120,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,4096,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,3072,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,2560,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,3584,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,2560,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,3584,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,3072,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,2048,0.004594666676388847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,3072,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,2560,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,1536,0.004021333323584663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,2560,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,2048,0.0070053330726093715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,1024,0.0033280000918441345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,2048,0.006364444477690591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,1536,0.006649777707126405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,768,0.0031742221779293488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,1536,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,512,0.002790222151411904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,1024,0.0063457778758472866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,768,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,256,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,6144,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,768,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,128,0.0022026666750510535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,512,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,256,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,64,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,256,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,32,0.0021280000607172647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,128,0.005762666463851929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,4096,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,65536,0.02882400155067444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,64,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,32,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,51200,0.023324444890022278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,65536,0.05674755573272705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,16384,0.012010667059156628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,65536,0.05605689022276136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,51200,0.045497778389188975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,12288,0.010414222048388587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,16384,0.016528000434239704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,51200,0.04448711209826999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,8192,0.011132444772455426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,10240,0.011372444530328115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,12288,0.014025777578353882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,8192,0.01037333326207267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,1024,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,10240,0.01275022245115704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,512,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,7168,0.009597332941161262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,10240,0.012483555409643384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,6144,0.008913777768611908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,8192,0.011056889262464313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,7168,0.010447111394670276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,8192,0.010789333118332757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,5120,0.008276444342401292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,7168,0.010072000324726105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,4096,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,128,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,6144,0.009434666898515489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,5120,0.009192888935407003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,3584,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,5120,0.008720888859695857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,4096,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,3584,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,4096,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,2560,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,3584,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,3072,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,2048,0.005484444399674733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,3072,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,2560,0.007439110842016008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,1536,0.004575110971927643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,16384,0.017665778597195942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,2560,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,2048,0.0069164443347189165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,2048,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,12288,0.014150222142537435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,1536,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,768,0.0032151111712058387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,1536,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,1024,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,512,0.002854222224818336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,1024,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,768,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,256,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,768,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,512,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,128,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,512,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,256,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,64,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,3072,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,256,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,32,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,128,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,65536,0.02345688806639777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,128,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,51200,0.018944000204404194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,64,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,32,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,65536,0.056291553709242076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,16384,0.012146666646003723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,65536,0.05459200011359321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,51200,0.044171555174721606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,12288,0.010335110955768162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,51200,0.04385066694683499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,16384,0.016700444949997794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,10240,0.009675555759006077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,16384,0.016859556237856548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,12288,0.013820444544156393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,8192,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,10240,0.012724444270133972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,12288,0.014147554834683737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,7168,0.00795999997191959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,8192,0.01106844428512785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,10240,0.012422222230169507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,6144,0.008184889124499427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,6144,0.010030222435792288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,7168,0.010399999717871347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,8192,0.010869333313571082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,5120,0.0074728892909155945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,4096,0.006801777829726537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,7168,0.010051555931568146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,6144,0.009676444033781687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,3584,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,5120,0.008973333570692275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,6144,0.009343110852771336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,3072,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,4096,0.008330666356616551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,5120,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,2560,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,3584,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,4096,0.008016889293988546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,2048,0.004381333374314838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,3584,0.007687110867765214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,2560,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,1536,0.0038648889296584656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,2048,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,2048,0.0063973334100511335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,768,1024,0.0035413333939181436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,1024,0.003185777821474605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,1536,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,768,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,1536,0.00601333338353369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,1024,0.006502222269773483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,512,0.002499555548032125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,768,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,1024,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,256,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,768,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,512,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,128,0.0021226666867733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,256,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,512,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,64,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,128,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,512,32,0.0018115554832749898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,256,0.004984000076850255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,65536,0.019370666808552213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,128,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,51200,0.01644355555375417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,64,0.004978666702906291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,65536,0.04948711064126757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,3072,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,32,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,16384,0.010747555229398938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,12288,0.00907911111911138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,65536,0.05401777889993456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,51200,0.03974666529231601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,2560,0.006702222343948152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,16384,0.016522667474216886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,10240,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,12288,0.013847110999955071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,16384,0.01687466601530711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,8192,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,10240,0.012543111211723752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,7168,0.00721955547730128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,10240,0.0123831108212471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,8192,0.011054221954610614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,6144,0.006647999915811751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,8192,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,7168,0.010327111515733931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,5120,0.007931555310885111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,7168,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,6144,0.009708444277445475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,4096,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,6144,0.009338666995366415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,5120,0.008976000050703684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,3584,0.006231110956933763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,5120,0.008620444271299574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,4096,0.008352888955010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,3072,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,4096,0.007818666597207388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,3584,0.00795999997191959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,2560,0.004863111095296012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,3584,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,3072,0.007244444555706448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,3072,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,51200,0.04335911075274149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,3072,0.00702044450574451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,1536,0.0036391110883818734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,2560,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,12288,0.013803555733627744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,2048,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,2048,0.006372444331645966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,1536,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,768,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,1024,0.006322666588756774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,1536,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,512,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,1024,0.005639111002286275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,768,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,256,0.0021964444054497615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,768,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,512,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,128,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,256,0.005856888989607493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,512,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,64,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,128,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,256,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,65536,0.02350666622320811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,128,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,64,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,2048,0.004238222208287981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,51200,0.020527111159430612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,65536,0.04927822285228305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,2560,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,16384,0.013642667068375481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,65536,0.051698668135537036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,12288,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,51200,0.03946311275164286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,1024,0.003156444471743372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,16384,0.016520000166363187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,51200,0.04132711225085788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,10240,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,16384,0.016493333710564505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,12288,0.01384800010257297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,12288,0.013777777552604675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,10240,0.012713777522246042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,7168,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,8192,0.011232888533009423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,10240,0.012424888710180918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,6144,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,8192,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,7168,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,5120,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,7168,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,6144,0.00963288876745436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,256,32,0.001871111078394784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,4096,0.006558222075303395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,6144,0.009330666727489894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,3584,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,5120,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,4096,0.00830577810605367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,3072,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,5120,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,3584,0.00816000004609426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,2560,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,32,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,2048,0.004233777936961916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,3072,0.007596444752481248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,3584,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,1536,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,2560,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,3072,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,1024,0.002891555635465516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,2560,0.006684444430801604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,1536,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,2048,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,768,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,8192,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,1536,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,1024,0.006320888797442119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,512,0.002524444419476721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,1024,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,768,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,256,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,768,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,128,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,256,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,256,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,64,0.0018604445374674266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,128,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,128,32,0.001812444378932317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,128,0.004871110949251386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,65536,0.01607644392384423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,64,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,4096,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,51200,0.013745778136783175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,32,0.005072000126043956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,16384,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,12288,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,65536,0.05066577924622429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,10240,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,51200,0.04091111156675551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,16384,0.016600888636377122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,8192,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,7168,0.005390222287840313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,12288,0.013557333085272046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,6144,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,10240,0.012095111111799875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,5120,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,512,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,4096,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,8192,0.010684444672531553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,512,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,7168,0.00981955561372969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,3584,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,3072,0.0052977779673205475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,6144,0.00903555585278405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,2560,0.0047031111187405055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,5120,0.00835111074977451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,2048,0.0042151109211974675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,1536,0.003543111185232798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,4096,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,3584,0.007366221812036302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,3072,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,768,0.0025439999169773525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,2048,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,512,0.002492444382773505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,1536,0.005867555737495422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,256,0.0021182222084866627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,1024,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,128,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,768,0.0052977779673205475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,2048,0.0069057775868309875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,64,0.0018115554832749898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,32,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,512,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,65536,0.015864888827006023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,256,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,128,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,64,0.004577777865860197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,16384,0.011216888825098673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,32,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,12288,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,10240,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,65536,0.05068177647060818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,8192,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,51200,0.04109422365824381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,7168,0.005305777821275923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,16384,0.016546666622161865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,12288,0.013872000078360239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,10240,0.012120889292822944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,64,1024,0.002854222224818336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,8192,0.01069777790043089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,4096,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,2560,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,7168,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,3584,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,3072,0.004951111144489712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,6144,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,2560,0.004703999807437261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,5120,0.00850044439236323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,4096,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,2048,0.0041946665280395085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,1536,0.0035048888789282907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,3584,0.007357333269384172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,1024,0.0028906667398081887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,3072,0.006809777683681912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,768,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,2560,0.006646222124497096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,512,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,51200,0.013590222431553734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,2048,0.006231110956933763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,256,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,1536,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,128,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,1024,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,64,0.0017946666727463405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,768,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,512,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,256,0.004871110949251386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,128,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,64,0.005017777697907554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,6144,0.005670222143332164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,32,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,5120,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,16384,0.3488462236192491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,16384,0.2755795584784614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,161,32,32,0.0018257778137922287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,51200,1.0933288998074002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,16384,0.1365200016233656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,12288,0.26704711384243435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,51200,0.40776801109313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,12288,0.21250666512383354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,12288,0.10352533393436009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,10240,0.24027466773986816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,51200,0.8493297894795736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,10240,0.17099911636776396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,10240,0.08611555231942071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,8192,0.19185422526465523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,16384,0.3073093361324734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,8192,0.1398693323135376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,7168,0.16754399405585396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,12288,0.2270115481482612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,8192,0.07112622261047363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,6144,0.14305244551764593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,7168,0.12328799565633138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,7168,0.06254666381412081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,10240,0.19373510943518746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,6144,0.10566133260726929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,5120,0.12291733423868816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,6144,0.055627557966444224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,8192,0.1542675495147705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,4096,0.1018951137860616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,5120,0.0874062246746487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,7168,0.13783821794721815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,3584,0.08671022123760647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,4096,0.07324888971116808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,6144,0.11787377463446723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,4096,0.03940711087650723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,5120,0.10076355271869236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,3072,0.07703822188907199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,4096,0.0837564468383789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,3584,0.0632773306634691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,3584,0.035303112533357404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,2560,0.06241244739956326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,3072,0.03131377696990967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,3584,0.07372622357474433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,2048,0.05066844489839342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,2560,0.04805066519313388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,3072,0.06556444697909884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,2560,0.05680711401833428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,5120,0.0465866658422682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,2048,0.023815999428431194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,2048,0.040977779361936785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,1536,0.03920088873969184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,2048,0.049588445160124034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,1536,0.03242222136921353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,1536,0.041165331999460854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,1024,0.0251875552866194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,768,0.02162844439347585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,1024,0.033977776765823364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,1024,0.015554666519165039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,3072,0.0558755530251397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,2560,0.027679999669392902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,512,0.015444444285498725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,768,0.025748444928063288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,51200,0.9135351181030273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,256,0.013415111435784234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,512,0.01789155602455139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,512,0.01330488920211792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,768,0.031207111146714952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,128,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,512,0.025603555970721777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,256,0.011767999993430244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,1024,0.028894222444958154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,64,0.007002666592597961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,1536,0.018788443671332467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,32,0.007165333463086023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,128,0.014675555957688225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,128,0.011536888778209686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,256,0.021260443660948012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,64,0.014454222387737699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,32,0.015309333801269531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,128,0.016798221402698092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,768,0.014118222726715935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,65536,0.8668213420444064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,256,0.020224889119466145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,51200,0.6680266592237684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,65536,1.1482017305162218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,51200,0.8995386759440104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,16384,0.29783911175198025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,16384,0.21569332811567518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,65536,0.4126364390055339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,12288,0.22180533409118652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,16384,0.10843911435869004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,10240,0.17004533608754477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,51200,0.3223048845926921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,12288,0.08202666706509061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,12288,0.16068177753024632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,10240,0.1381093396080865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,10240,0.06930133369233873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,8192,0.13847022586398655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,12288,0.18562400341033936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,16384,0.24299200375874838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,7168,0.12220356199476455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,8192,0.11345599757300483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,8192,0.056919111145867236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,10240,0.15481066703796387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,7168,0.09718044598897298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,7168,0.050606220960617065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,8192,0.12559378147125244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,6144,0.08610933356814915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,7168,0.1124417781829834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,5120,0.08956444263458252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,6144,0.043868445687823825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,6144,0.09711733129289414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,5120,0.07039022445678711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,5120,0.03776889046033224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,4096,0.0762337777349684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,4096,0.03256977929009332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,5120,0.08308888806237115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,4096,0.058111998769972056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,3584,0.06637066602706909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,6144,0.10444888803693984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,3072,0.056672890981038414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,4096,0.07024977604548137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,3584,0.051631109582053296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,3584,0.02903377678659227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,2560,0.04808800088034736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,3072,0.026212443908055622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,3072,0.04502400093608432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,3584,0.06361422273847792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,2048,0.04004888733228048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,2560,0.02351288828584883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,3072,0.056440889835357666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,1536,0.031009776724709406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,2560,0.050266666544808276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,2048,0.03358577688535055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,2048,0.043791999419530235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,1536,0.027143999934196472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,1024,0.023167999254332647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,1536,0.03659111261367798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,1024,0.020967110991477966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,768,0.017488888568348356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,1024,0.01389600005414751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,51200,0.7601164182027181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,768,0.019188443819681805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,1024,0.030598223209381104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,512,0.012164444559150271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,768,0.012647999657524956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,768,0.02608088817861345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,256,0.008385777473449707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,2048,0.019748444358507793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,512,0.015295111470752291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,512,0.011583111352390714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,128,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,512,0.022636445032225713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,256,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,1536,0.016505777835845947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,256,0.013445333474212222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,64,0.005433777968088786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,32,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,128,0.010482666393121084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,256,0.01926577753490872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,65536,1.0066613091362846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,128,0.012494222157531314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,128,0.014783110883500842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,32,0.013394667042626275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,2560,0.039346665143966675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,65536,0.3037217722998725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,51200,0.2980835437774658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,16384,0.09708178043365479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,65536,0.14143733183542886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,51200,0.11139733261532253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,16384,0.07928622431225248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,12288,0.0759342246585422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,16384,0.04003377755482992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,64,0.012400888734393649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,12288,0.059848891364203564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,12288,0.03184444374508328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,65536,0.3792204327053494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,10240,0.0639244450463189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,10240,0.027671999401516382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,12288,0.06900622447331746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,51200,0.23687556054857042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,8192,0.052301334010230176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,51200,0.25378666983710396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,7168,0.04609333475430807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,65536,0.3288399908277723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,8192,0.023356444305843774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,10240,0.060028447045220264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,8192,0.04359022114011976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,6144,0.039579557047949895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,7168,0.021359110871950786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,7168,0.03805955582194858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,8192,0.05105244451098972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,5120,0.0339902208911048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,16384,0.08636089166005452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,6144,0.018761777215533786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,6144,0.03354133168856303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,7168,0.04571733209821913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,4096,0.02753244505988227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,5120,0.028806222809685603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,6144,0.04112177756097581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,3584,0.024676443801985845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,5120,0.03598755598068237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,4096,0.014861333701345654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,4096,0.024350222614076402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,3072,0.022016889519161646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,10240,0.05149155524041918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,3584,0.013783999615245394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,3584,0.02198933396074507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,4096,0.032167110178205706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,2560,0.01825066738658481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,3584,0.028978665669759113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,3072,0.019343111250135634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,3072,0.012770666844315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,2048,0.015159999330838522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,2560,0.016898666818936665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,3072,0.02717155549261305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,2560,0.011914666328165265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,1536,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,2560,0.024439111351966858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,2048,0.010864888628323873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,2048,0.014854222536087036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,1024,0.009330666727489894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,1536,0.008683555656009251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,1536,0.0127368892232577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,2048,0.02188355558448368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,5120,0.016743110285864938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,768,0.007728888756699032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,1024,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,1024,0.00790666706032223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,1536,0.01920622256067064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,512,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,768,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,1024,0.016890666551060148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,256,0.0041911109454101985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,512,0.007251555720965068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,768,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,512,0.008254222571849823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,768,0.016964443855815463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,256,0.007356444166766272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,512,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,64,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,128,0.007301333049933116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,128,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,32,0.0031679999083280563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,64,0.007330666813585494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,32,0.007888889147175683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,65536,0.2757733398013645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,65536,0.21954754988352457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,51200,0.21446666452619764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,128,0.0035235554807715942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,65536,0.12600978215535483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,256,0.00675022229552269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,16384,0.07358577516343859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,256,0.011427555647161273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,51200,0.171195560031467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,51200,0.10061955451965332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,12288,0.055536889367633395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,16384,0.037189334630966187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,16384,0.08546133173836602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,12288,0.029750221305423315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,12288,0.044255110952589244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,10240,0.04661155409283108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,12288,0.06769955820507474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,65536,0.3058195643954807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,10240,0.02548533346917894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,8192,0.039552887280782066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,51200,0.24125422371758354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,7168,0.03402400016784668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,8192,0.02164977788925171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,10240,0.05855288770463732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,8192,0.03278844555219015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,6144,0.02937777837117513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,7168,0.019696000549528334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,8192,0.04979911115434435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,7168,0.028664890262815688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,16384,0.05832266807556152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,5120,0.025376889440748427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,128,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,6144,0.02533600065443251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,7168,0.04501422246297201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,4096,0.0214008887608846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,5120,0.015953777564896476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,6144,0.04059999850061204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,3584,0.01867555578549703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,4096,0.014496000276671516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,5120,0.03553244471549988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,10240,0.038247111770841814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,3072,0.016515556308958266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,4096,0.03095910946528117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,3584,0.013221333424250284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,3584,0.016487111647923786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,2560,0.013764444324705334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,3072,0.012481778032249875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,3584,0.028198222319285076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,3072,0.015360888507631091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,2560,0.01369066701995002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,2560,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,3072,0.025966222087542217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,2048,0.01178311142656538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,6144,0.017893332574102614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,1536,0.009773333039548662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,2048,0.009659555223253038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,5120,0.021752888957659405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,2048,0.01239466667175293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,1024,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,1536,0.01072533345884747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,1536,0.008344000412358178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,1024,0.00872444444232517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,1024,0.007663999994595845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,768,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,1024,0.01550044450494978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,768,0.008344000412358178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,512,0.004837333328194088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,768,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,768,0.014820444915029736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,512,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,512,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,256,0.0035546666218174826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,2560,0.02332266668478648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,512,0.013366222381591797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,128,0.0031991110493739447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,2048,0.021239999267790053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,256,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,4096,0.018577777677112155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,64,0.0032097777972618737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,32,0.0031759999692440033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,128,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,1536,0.018944889307022095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,64,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,128,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,32,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,65536,0.2439333332909478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,65536,0.20347466733720568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,65536,0.12310755252838135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,51200,0.15955289204915366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,16384,0.06430400080151029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,51200,0.09816710816489325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,256,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,16384,0.055108444558249585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,256,0.011524444652928246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,128,0.00666222224632899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,12288,0.05008088880115085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,16384,0.03624533282385932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,10240,0.04171289006868998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,16384,0.08453689018885295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,12288,0.028468443287743464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,12288,0.04222133424546984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,65536,0.3051075670454237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,51200,0.23989333046807182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,8192,0.035517334938049316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,51200,0.19661333825853136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,10240,0.03640800052218967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,10240,0.02437866727511088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,8192,0.0211182220114602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,8192,0.031129777431488037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,6144,0.02699111070897844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,7168,0.019267555740144517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,7168,0.027330666780471802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,8192,0.04917066627078586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,5120,0.023256889647907678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,6144,0.023863111933072407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,7168,0.0440577765305837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,5120,0.015761777758598328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,6144,0.03973066806793213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,5120,0.020591111646758187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,4096,0.019842666056421068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,4096,0.014136888914638095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,5120,0.03471555643611484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,3584,0.016930666234758165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,4096,0.01776622235774994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,7168,0.030760000149408977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,3072,0.015096000499195524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,3584,0.013072888884279462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,3584,0.016238222519556682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,10240,0.05779733260472616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,3584,0.027219555444187585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,3072,0.01202400028705597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,3072,0.014937778313954672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,6144,0.01737422247727712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,2560,0.010626667075686984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,2048,0.011101333631409539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,2560,0.013493333425786761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,2048,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,2560,0.02313599983851115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,1536,0.009213333328564962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,2048,0.012055111428101858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,1536,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,2048,0.02090488870938619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,1024,0.007530666887760162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,1536,0.010321777727868822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,12288,0.06657066610124376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,768,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,4096,0.03002133303218418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,1536,0.018212444252438016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,1024,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,1024,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,512,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,768,0.007243555453088548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,2560,0.012753778033786349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,768,0.007838221887747446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,1024,0.014815110299322339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,256,0.003830222205983268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,512,0.007066667079925537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,768,0.014440889159838358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,3072,0.025212445192866858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,512,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,128,0.0031600000543726813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,256,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,256,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,64,0.002846222163902389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,128,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,512,0.013351110948456658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,128,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,65536,0.17954133616553414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,128,0.009425777528021071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,32,0.006992888947327931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,51200,0.14061778121524388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,65536,0.1622426642311944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,65536,0.07574666870964898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,51200,0.12715910540686712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,16384,0.04828800095452202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,51200,0.060454222891065806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,32,0.0029964444951878656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,16384,0.04462755719820658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,65536,0.17017243968115914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,12288,0.03882577684190538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,16384,0.023757333556811016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,64,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,10240,0.03173866536882188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,16384,0.05060888992415535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,12288,0.01891999940077464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,10240,0.029481778542200725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,10240,0.01679022279050615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,12288,0.04139111108250088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,8192,0.02696177860101064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,10240,0.036269333627488874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,7168,0.023224888576401606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,8192,0.024833778540293377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,8192,0.031938665442996554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,7168,0.013737777868906656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,6144,0.020376000139448378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,7168,0.021918222308158875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,256,0.0110880004035102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,5120,0.017489777670966256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,51200,0.13539821571773952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,7168,0.028825776444541082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,6144,0.012787555654843649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,6144,0.019575110740131803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,4096,0.014475555883513557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,5120,0.011756444142924415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,12288,0.034435556994544134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,6144,0.026362667481104534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,3584,0.012700444294346703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,4096,0.015127999915017022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,5120,0.023265777362717524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,3072,0.011354666617181567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,3584,0.013925333817799887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,3584,0.00965244405799442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,4096,0.02105866703722212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,8192,0.01537333263291253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,2560,0.00979288915793101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,3072,0.008906666603353288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,3584,0.02000177734427982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,3072,0.012738666600651212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,2048,0.008380444513426887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,2560,0.008220444122950235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,2560,0.011566222541862063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,1536,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,3072,0.018555555078718398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,2048,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,2048,0.010629333555698395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,2560,0.017105778058369953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,1024,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,1536,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,2048,0.016352888610627916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,1536,0.008773333496517604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,768,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,5120,0.01721600029203627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,1024,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,1536,0.014256000518798828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,512,0.00378577783703804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,4096,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,768,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,768,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,1024,0.012459555433856117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,256,0.002977777686383989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,512,0.006303999986913469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,768,0.012306667036480375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,128,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,256,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,256,0.006043555421961679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,64,0.002463111033042272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,512,0.0111191107167138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,128,0.006057777752478917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,32,0.00250133333934678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,128,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,128,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,65536,0.16533154911465114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,32,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,1024,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,65536,0.15709955162472194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,65536,0.07182044453091092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,51200,0.13696355289883083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,512,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,16384,0.045102222098244556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,51200,0.12314400408003066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,51200,0.058196445306142174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,65536,0.15864977571699354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,256,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,16384,0.022680888573328655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,12288,0.03547022077772353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,16384,0.04314844475852119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,10240,0.029488888051774766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,64,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,51200,0.12527910868326822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,12288,0.01860000027550591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,8192,0.025081778566042583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,10240,0.02877955635388692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,12288,0.03980355461438497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,7168,0.021890666749742296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,8192,0.02437777817249298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,8192,0.01456622282663981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,7168,0.013439111411571503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,8192,0.03152711192766825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,6144,0.018945778409639995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,7168,0.028082665469911363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,6144,0.01923822197649214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,5120,0.016234666109085083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,6144,0.012416889270146688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,6144,0.02570666703912947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,12288,0.03355911042955186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,4096,0.013745778136783175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,5120,0.01689511040846507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,5120,0.011543110840850405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,10240,0.01644799941115909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,3584,0.012063999970753988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,4096,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,5120,0.023226666781637404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,10240,0.03570133447647095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,3072,0.011003555523024665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,3584,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,3584,0.013656888571050433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,4096,0.020827554994159274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,7168,0.02120888895458645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,2560,0.009322666459613377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,3072,0.008347555167145198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,3072,0.012723555167516073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,3584,0.019711110326978896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,2048,0.008260444634490544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,2560,0.011375111010339526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,3072,0.018571555614471436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,1536,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,2048,0.01035288886891471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,2560,0.017198221551047433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,2048,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,1024,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,16384,0.050287110937966235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,1536,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,1536,0.00850133349498113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,768,0.00453599997692638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,1024,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,1024,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,1536,0.013293333351612091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,512,0.003699555579158995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,4096,0.014767110347747803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,768,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,1024,0.011743110915025076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,768,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,256,0.0029279999434947968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,512,0.006123555617200003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,512,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,768,0.012132444315486483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,128,0.0028844444702068963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,256,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,256,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,512,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,64,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,2560,0.007754666937722101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,32,0.0026417778184016547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,256,0.009716444545321995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,128,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,65536,0.14693955580393472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,128,0.007729777859316931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,32,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,51200,0.12213955985175239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,65536,0.11956889099544948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,65536,0.06988266441557142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,16384,0.046032889021767505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,51200,0.09299200110965306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,51200,0.056419557995266385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,65536,0.14680355125003391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,12288,0.032660444577534996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,16384,0.0339937773015764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,51200,0.11728978157043457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,128,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,16384,0.046979556481043495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,10240,0.027377777629428442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,12288,0.017813333206706576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,64,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,12288,0.03842577669355605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,2048,0.015843555331230164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,8192,0.023349333140585158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,10240,0.016134222348531086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,10240,0.022367111510700647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,7168,0.020271110865804885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,8192,0.014022222823566861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,8192,0.019154666198624503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,10240,0.03333955672052171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,6144,0.017514665921529133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,7168,0.017086222767829895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,8192,0.028718223174413044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,16384,0.021915555000305176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,5120,0.01519466605451372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,6144,0.015621332658661736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,6144,0.012074666718641916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,7168,0.026336888472239178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,4096,0.012482666307025485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,5120,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,12288,0.026015111141734656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,6144,0.024494222468800012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,5120,0.013915555344687568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,3584,0.011392888923486074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,4096,0.012617777619096967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,5120,0.02188088827662998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,3072,0.010314666562610203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,3584,0.008752000000741746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,3584,0.011519999967681037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,4096,0.01993333299954732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,2560,0.009578666753239101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,3584,0.0188426673412323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,3072,0.011028444601429833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,7168,0.013241777817408243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,2560,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,2048,0.008361777497662438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,3072,0.01752711170249515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,2048,0.0073315559162033936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,1536,0.006870222174459034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,2048,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,2048,0.014879999889267815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,1536,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,4096,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,1024,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,1536,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,768,0.004423999951945411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,1536,0.013099555340078143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,1024,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,1024,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,512,0.0042257776690853965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,3072,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,768,0.006736889067623351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,768,0.006314666734801398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,256,0.0032088889016045462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,1024,0.011720000041855706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,512,0.006193777753247156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,2560,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,768,0.011391999820868174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,128,0.00281866660548581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,256,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,256,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,64,0.00254133323000537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,512,0.010337777435779572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,2560,0.016178665889634024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,32,0.002550222186578645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,128,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,256,0.009986666341622671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,65536,0.13792178365919325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,64,0.00563822231358952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,32,0.005889777921968036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,51200,0.12813511159684923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,65536,0.11612710687849258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,65536,0.06684355603324042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,16384,0.04574577675925361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,51200,0.09020355674955581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,51200,0.05420710974269443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,65536,0.14603200223710802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,12288,0.034838222795062594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,16384,0.020964443683624268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,16384,0.033118221494886614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,512,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,51200,0.11662933561537002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,10240,0.029286222325430974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,12288,0.01721333298418257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,16384,0.046241776810752026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,8192,0.02460799945725335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,10240,0.02176888949341244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,10240,0.015494222442309061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,128,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,7168,0.021469333105617102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,8192,0.013752889302041797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,10240,0.03250044584274292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,8192,0.018587556150224473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,6144,0.018389332625601027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,7168,0.012748444245921241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,8192,0.028340445624457464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,7168,0.016882666283183627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,6144,0.011699555648697747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,7168,0.026139555705918208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,5120,0.01590133375591702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,6144,0.01514933341079288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,6144,0.023982221881548565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,4096,0.013506666653686099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,5120,0.010312889185216691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,5120,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,12288,0.025012445118692186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,3584,0.011798222031858234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,5120,0.02159822152720557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,4096,0.01236355553070704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,12288,0.03734577695528666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,3072,0.010748444332016839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,3584,0.008372444245550368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,4096,0.01964977714750502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,3584,0.01144266708029641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,2560,0.00908088849650489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,3072,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,3584,0.0181048893266254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,3072,0.01033866653839747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,2048,0.008172444171375698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,2560,0.00909511082702213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,2560,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,3072,0.01717688971095615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,1536,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,2048,0.007287110719415877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,2048,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,2560,0.015781333049138386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,1024,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,1536,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,1536,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,768,0.004228444563017951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,128,0.005631111148330901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,2048,0.014902222487661572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,1024,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,4096,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,512,0.0035084444615576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,1024,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,1536,0.012426666915416718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,768,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,256,0.0031617778456873367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,768,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,512,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,1024,0.011247110863526663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,128,0.002560888934466574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,512,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,256,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,768,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,256,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,32,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,128,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,128,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,256,0.009026666482289633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,65536,0.0992106662856208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,64,0.005670222143332164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,128,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,32,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,51200,0.07709244224760267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,65536,0.10830755366219415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,65536,0.06463199853897095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,51200,0.05179466803868612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,65536,0.14476888709598118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,16384,0.03136444422933791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,12288,0.025952888859642878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,51200,0.11592000060611302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,16384,0.020336000455750358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,16384,0.045648889409171216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,12288,0.016843555702103507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,12288,0.024244444237815008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,512,0.009992000129487779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,12288,0.03695199886957804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,10240,0.01515111161602868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,8192,0.01622488929165734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,10240,0.02091466718249851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,8192,0.013280889226330651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,10240,0.032046222024493747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,7168,0.01407911048995124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,8192,0.01806311143769158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,51200,0.08580444256464641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,6144,0.012413333687517377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,7168,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,8192,0.028042667441897925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,7168,0.016506666938463848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,5120,0.011085333095656501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,6144,0.010965333216720156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,7168,0.02516888909869724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,6144,0.015095111396577625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,4096,0.01092888911565145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,5120,0.013431111143694984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,5120,0.009657777845859528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,64,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,6144,0.023136888941129048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,10240,0.018876444962289598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,3584,0.010009778042634329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,4096,0.00852177788813909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,4096,0.011687999798191918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,3072,0.007585778004593319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,3584,0.00831911133395301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,5120,0.02130844526820713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,3584,0.011019555230935415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,2560,0.006562666760550604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,4096,0.019127999742825825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,3072,0.009588444398509132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,3584,0.01788444485929277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,16384,0.031565331750445895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,2048,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,3072,0.01684533390733931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,2048,0.007900444169839224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,2560,0.015097777048746744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,1536,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,2048,0.013779555757840475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,1536,0.007297777467303806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,1536,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,1024,0.003548444559176763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,1536,0.012092444631788464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,1024,0.006446222050322427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,1024,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,3072,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,1024,0.011053333679835001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,768,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,768,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,2560,0.008906666603353288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,2048,0.00703999979628457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,768,0.011034666664070554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,256,0.002546666603949335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,512,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,512,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,128,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,256,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,256,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,128,0.0053004444473319584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,256,0.009308444129096137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,128,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,64,0.002159111201763153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,32,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,768,0.0031662223239739737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,65536,0.09370222356584336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,128,0.007666666474607255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,64,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,32,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,51200,0.07401866383022733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,65536,0.06336889002058241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,2560,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,65536,0.10569066471523708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,16384,0.029362665282355413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,51200,0.08357777860429551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,51200,0.05083733465936449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,12288,0.02327288852797614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,65536,0.14536800649431017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,16384,0.030527111556794908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,16384,0.019906666543748643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,10240,0.020038222273190815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,51200,0.11509244971805149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,512,0.00999111102686988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,12288,0.016511110795868766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,16384,0.04530844423505995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,8192,0.015084443820847405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,10240,0.015085332923465304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,10240,0.020618667205174763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,7168,0.01371022231049008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,12288,0.03586488962173462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,8192,0.012715555727481842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,8192,0.017681777477264404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,10240,0.03133422136306763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,6144,0.011960888902346293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,7168,0.011354666617181567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,7168,0.015984889533784654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,8192,0.027552000350422327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,5120,0.01070755554570092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,6144,0.01035288886891471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,512,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,6144,0.014644443988800049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,4096,0.011355555719799466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,5120,0.009312888814343346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,6144,0.023357333408461675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,3584,0.009679111341635386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,4096,0.011377777490350934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,4096,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,5120,0.021341333786646526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,3072,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,4096,0.019102222389645047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,3584,0.008231999973456064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,3584,0.010604444477293225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,12288,0.023652444283167522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,2560,0.006690666493442323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,3072,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,3072,0.009975999593734741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,2048,0.005658666706747479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,2560,0.007612444460391998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,2560,0.008345777789751688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,3072,0.016007110476493835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,1536,0.0046826667255825466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,2048,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,2048,0.007232888705200619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,7168,0.025432000557581585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,1536,0.0069884442620807225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,2048,0.013328889177905189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,1536,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,5120,0.01330044451687071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,1536,0.012768000364303589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,1024,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,768,0.0031831111345026228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,1024,0.011028444601429833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,768,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,512,0.002862222285734283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,3584,0.016925333274735343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,768,0.011086222198274402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,512,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,512,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,256,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,2560,0.01481599940194024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,512,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,1024,0.0038533334930737815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,256,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,256,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,1024,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,128,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,64,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,32,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,128,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,128,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,65536,0.07941777838601007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,64,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,32,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,51200,0.06165777974658542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,65536,0.06743911239835951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,65536,0.14452977975209555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,128,0.002209777840309673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,256,0.005882666756709416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,16384,0.02465244465404087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,51200,0.049714667929543384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,51200,0.11514666345384385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,12288,0.020894222789340552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,16384,0.020679111282030743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,16384,0.01973511113060845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,10240,0.017837334010336135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,16384,0.044180442889531456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,12288,0.01668711172209846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,12288,0.016506666938463848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,8192,0.01754399968518151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,10240,0.014760888285107084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,10240,0.014695111248228284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,12288,0.035868446032206215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,7168,0.015370666980743408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,65536,0.06265244219038221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,8192,0.013057777451144325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,8192,0.0124177775449223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,6144,0.01421511173248291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,10240,0.031327111853493586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,51200,0.05327555537223816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,7168,0.011839110818174152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,5120,0.01202755586968528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,8192,0.027282666828897264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,6144,0.010031999813185798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,6144,0.011044444309340583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,7168,0.025055999557177227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,4096,0.010139555566840703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,5120,0.009313777916961247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,5120,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,6144,0.02295111119747162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,3584,0.008774222599135505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,4096,0.008367111285527548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,4096,0.00849777791235182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,3072,0.008033778104517195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,3584,0.00829777783817715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,3584,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,768,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,2560,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,3584,0.017234666479958426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,3072,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,3072,0.007663999994595845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,2048,0.006230222268237009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,2560,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,3072,0.01605422298113505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,7168,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,1536,0.00462755560874939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,2560,0.014826666977670459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,2048,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,1536,0.007027555671003129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,2048,0.013520888984203339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,1536,0.00630666646692488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,1024,0.003823111040724648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,1024,0.006411555740568373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,5120,0.021008888880411785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,768,0.0032764443506797156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,1024,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,4096,0.018912000788582694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,512,0.0028320000403457214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,1024,0.011083555718262991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,768,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,512,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,512,0.005438222239414851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,2560,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,256,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,256,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,128,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,256,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,256,0.009142222503821055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,128,0.004952888935804367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,1536,0.0127368892232577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,128,0.005714666512277391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,128,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,64,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,32,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,65536,0.06666755676269531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,768,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,64,0.005159111072619756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,32,0.005395555661784277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,51200,0.05296711126963297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,768,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,65536,0.06508800056245592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,65536,0.0610062215063307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,16384,0.022103110949198406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,2048,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,51200,0.05099822084108988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,51200,0.048680000834994845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,65536,0.14424888292948404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,12288,0.01828533411026001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,16384,0.019959999455346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,51200,0.11509422461191814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,10240,0.01590844492117564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,12288,0.015843555331230164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,12288,0.016188444362746347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,12288,0.0353999998834398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,10240,0.014096000128322177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,8192,0.016340444485346477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,10240,0.01441422270403968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,8192,0.012014221813943652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,10240,0.03126044405831231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,7168,0.014427555931939019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,8192,0.01260088880856832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,7168,0.01072799993885888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,8192,0.02715377840730879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,512,0.010019555687904358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,7168,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,7168,0.025126222107145522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,6144,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,6144,0.010671999719407825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,6144,0.023016888234350417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,16384,0.043882668018341064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,4096,0.009679111341635386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,5120,0.00923555592695872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,5120,0.020682666036817763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,3584,0.008894222478071848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,4096,0.00830488900343577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,4096,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,4096,0.01761866609255473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,3584,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,3072,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,3584,0.007919111185603673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,6144,0.013005333642164866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,2560,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,3072,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,5120,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,3072,0.015501333607567681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,2048,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,2560,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,2560,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,5120,0.009395555489593083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,1536,0.004583999928500918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,2048,0.007276444799370236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,2048,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,2048,0.01346933344999949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,1024,0.003573333223660787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,1536,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,1536,0.012323555847009024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,1024,0.005713777823580637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,768,0.0031973332580592898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,1024,0.006593777901596493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,3072,0.007703999678293864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,1024,0.010743111371994019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,512,0.0028577778074476454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,768,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,16384,0.019345778557989333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,768,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,256,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,512,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,512,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,768,0.010695999695195092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,2560,0.014138667119873894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,128,0.0023128889087173673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,256,0.005727111051479976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,256,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,512,0.00997688869635264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,64,0.0021262222694026097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,32,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,128,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,1536,0.006649777707126405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,128,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,256,0.00980799976322386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,65536,0.054133332437939115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,64,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,128,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,32,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,65536,0.059671110577053495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,65536,0.062136001057095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,16384,0.018371555540296767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,3584,0.016532444291644625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,51200,0.04731288883421156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,51200,0.04931555522812737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,12288,0.014576888746685453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,65536,0.14399022526211208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,16384,0.01903288894229465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,51200,0.11509777439965142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,10240,0.013022222452693515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,12288,0.015487111277050443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,16384,0.0440942214594947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,12288,0.015840889679061044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,8192,0.0129439996348487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,10240,0.01352266636159685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,12288,0.034825778669781156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,10240,0.014103111293580798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,7168,0.01165244479974111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,8192,0.012099555797047086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,10240,0.031015111340416804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,51200,0.0442737771405114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,6144,0.010791111323568555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,7168,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,7168,0.011392888923486074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,8192,0.02680444386270311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,5120,0.010248888697889116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,6144,0.01001422190003925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,7168,0.024734222226672705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,6144,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,4096,0.008940444224410588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,16384,0.0192266669538286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,5120,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,5120,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,6144,0.022316444251272414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,3584,0.007920888562997183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,4096,0.008645333349704742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,4096,0.008225777910815345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,5120,0.020577778418858845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,3072,0.007391110890441471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,3584,0.008316444853941599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,3584,0.007707555260923173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,2560,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,4096,0.017320889565679762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,3072,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,3584,0.0162426663769616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,2048,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,3072,0.007276444799370236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,2560,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,2560,0.0069057775868309875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,1536,0.0041911109454101985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,3072,0.01521511044767168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,2048,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,2560,0.014404444230927361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,1024,0.0035377778112888336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,1536,0.006727111008432176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,1536,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,2048,0.013083555632167392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,768,0.0031555555760860443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,1024,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,1024,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,512,0.002826666666401757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,768,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,1024,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,768,0.010644444160991244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,512,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,256,0.002463999928699599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,512,0.009977777798970541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,8192,0.011436444189813403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,256,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,128,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,256,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,2048,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,256,0.009654222263230218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,64,0.002143111079931259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,32,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,128,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,65536,0.04381066560745239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,64,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,128,0.007419555551475949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,1536,0.012062222593360476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,32,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,51200,0.03513066636191474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,768,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,65536,0.059671110577053495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,65536,0.058077335357666016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,16384,0.01479644411140018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,51200,0.04708622230423821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,51200,0.046578665574391685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,12288,0.012682666381200155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,16384,0.0188391109307607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,16384,0.018265777164035372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,51200,0.11424977249569362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,10240,0.01262399968173769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,16384,0.043750223186280995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,12288,0.01479555500878228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,12288,0.01538133290078905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,8192,0.012691555751694573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,10240,0.013059555656380124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,10240,0.013532444834709167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,128,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,12288,0.03476266728507148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,7168,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,8192,0.011438222395049201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,8192,0.011342222491900126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,10240,0.030784888399971858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,6144,0.010185778141021729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,7168,0.01074488874938753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,7168,0.010579555398888059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,8192,0.026850667264726426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,5120,0.008972444468074376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,6144,0.009711110757456886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,7168,0.023778667052586872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,6144,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,512,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,4096,0.007673777639865875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,5120,0.009033777647548253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,65536,0.14334132936265734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,5120,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,3584,0.006908444480763541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,4096,0.008640888664457533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,4096,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,3072,0.0063893331421746155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,5120,0.01917155583699544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,3584,0.008268444074524773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,3584,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,2560,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,3072,0.007899555895063613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,3584,0.01640888883007897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,2560,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,2560,0.007355555891990662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,2048,0.004547555413511065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,2560,0.014147554834683737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,2048,0.00656088896923595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,1536,0.00388711111413108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,2048,0.007089777953094906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,2048,0.013064888616402945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,6144,0.021533333592944678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,1024,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,1536,0.006015111174848344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,1536,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,4096,0.017608000172509086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,1536,0.011714666253990598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,768,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,1024,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,3072,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,512,0.002521777732504739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,768,0.006074666563007567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,1024,0.011040888726711273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,768,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,3072,0.015107555521859063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,512,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,768,0.010743111371994019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,512,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,128,0.0021066665649414062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,256,0.005737777799367905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,512,0.009764444496896531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,64,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,128,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,128,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,256,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,32,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,65536,0.0317537784576416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,64,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,128,0.0073626670572492815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,51200,0.02641777859793769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,32,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,1024,0.0063697778516345555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,65536,0.057004445128970675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,65536,0.05714844332800972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,16384,0.012118221984969245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,51200,0.04608444372812907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,51200,0.04531466629770067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,12288,0.012651555240154266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,65536,0.14311733510759142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,16384,0.01827111177974277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,256,0.002303999952144093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,16384,0.01683288812637329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,10240,0.011587555209795633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,51200,0.11363466580708821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,12288,0.01406311160988278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,16384,0.04367644588152567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,12288,0.014198222094111972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,8192,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,256,0.004959110998445087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,7168,0.009633777870072259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,12288,0.03367644548416138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,8192,0.01106844428512785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,8192,0.011143999795118967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,10240,0.030357334348890517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,6144,0.008720888859695857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,7168,0.010372444159454769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,7168,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,8192,0.02575911084810893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,5120,0.008174222376611497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,6144,0.009683555199040307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,6144,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,7168,0.023365333676338196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,4096,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,5120,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,5120,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,6144,0.021390222840838965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,3584,0.010661332971519895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,4096,0.00831288927131229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,5120,0.018872888551818002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,4096,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,3072,0.009186666872766282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,3584,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,3584,0.016171556380059984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,10240,0.012738666600651212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,2560,0.008071111308203803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,3072,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,3072,0.014797333214018079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,2560,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,2048,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,2560,0.014143110977278816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,2048,0.006362666686375936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,2048,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,1536,0.0041457778877682155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,3584,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,2048,0.013112000293201871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,10240,0.0127697777416971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,1024,0.0035146665241983202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,1536,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,1536,0.012076444096035428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,1024,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,768,0.00313688897424274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,2560,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,1024,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,768,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,512,0.0027928888383838865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,768,0.010064889159467485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,4096,0.016940444707870483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,512,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,256,0.0024604443460702896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,512,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,3072,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,128,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,1536,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,256,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,256,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,64,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,128,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,32,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,256,0.009323555562231276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,128,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,65536,0.024340444140964087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,1024,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,128,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,64,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,51200,0.02199733257293701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,65536,0.05693333016501533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,65536,0.056234664387173124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,768,0.005988444305128521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,16384,0.012512889173295764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,51200,0.044580443037880786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,51200,0.045758220884535045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,65536,0.14278311199612087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,12288,0.0107004443804423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,16384,0.017441777719391715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,16384,0.016716443830066256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,10240,0.011775111158688864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,51200,0.11379822095235188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,12288,0.013950222068362765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,16384,0.04279111160172356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,12288,0.014146667387750415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,8192,0.010403555300500657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,512,0.009992000129487779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,10240,0.012844444149070315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,12288,0.034592890077167086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,7168,0.009718221922715506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,8192,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,8192,0.010796444283591377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,10240,0.029692444536421034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,6144,0.008808000220192803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,7168,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,8192,0.025386666258176167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,32,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,6144,0.009325332939624786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,6144,0.010008888940016428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,5120,0.009039999710188972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,4096,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,5120,0.019310222731696237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,4096,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,10240,0.012606222596433429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,4096,0.008649778034951951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,3584,0.00674933319290479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,4096,0.01722133325205909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,3584,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,7168,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,3072,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,5120,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,3584,0.015847111741701763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,7168,0.0233608881632487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,2560,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,3072,0.007916444705592262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,3072,0.007247111035717859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,2048,0.004508444418509801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,6144,0.02125066684352027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,2560,0.0075786668393347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,2560,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,3072,0.015484443969196744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,1536,0.0038764443662431505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,2048,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,2048,0.006513777706358168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,2560,0.014076444837782117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,1024,0.0032008888406885993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,1536,0.006106666806671355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,2048,0.013083555632167392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,1536,0.011752888560295105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,1024,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,1024,0.005774222314357758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,768,0.0028568889117903183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,768,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,512,0.002528888897763358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,1024,0.011375111010339526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,3584,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,768,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,256,0.0022159999029503926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,512,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,768,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,512,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,5120,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,128,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,256,0.005099555684460534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,256,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,64,0.0019288888821999233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,512,0.01036711119943195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,768,32,0.0018648888087934917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,128,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,128,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,65536,0.025030222203996446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,256,0.009659555223253038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,64,0.004967111266321606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,128,0.007584888901975419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,32,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,65536,0.0564240018526713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,65536,0.05449955662091573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,16384,0.014427555931939019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,51200,0.044128888183169894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,51200,0.04404533240530226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,65536,0.14244177606370714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,12288,0.01034577770365609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,16384,0.016502221425374348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,16384,0.016861332787407767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,51200,0.11325866646236843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,10240,0.009415110780133141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,12288,0.01390044391155243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,12288,0.014108444253603617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,16384,0.043282667795817055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,8192,0.007948444121413762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,10240,0.012473777764373355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,10240,0.012720000412729053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,12288,0.03342399994532267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,7168,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,8192,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,8192,0.01091288857989841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,10240,0.029291556941138372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,1536,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,51200,0.021499555971887376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,7168,0.010519111322032081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,7168,0.01018133345577452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,8192,0.025496888491842482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,6144,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,6144,0.009368889033794403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,7168,0.02278577784697215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,4096,0.006505777852402792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,6144,0.020976000361972388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,5120,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,3584,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,4096,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,4096,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,5120,0.019371555911170114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,3072,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,4096,0.01719111038578881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,3584,0.00795199970404307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,3072,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,2560,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,3584,0.015861334072219003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,6144,0.007184888753626082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,3072,0.01514755520555708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,2048,0.004361777669853634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,5120,0.00683822234471639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,2560,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,2560,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,1536,0.0038542221817705366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,2048,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,2560,0.014143110977278816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,2048,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,5120,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,1024,0.0031591111587153543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,1536,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,2048,0.012747555143303342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,768,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,1024,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,1024,0.00564444437623024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,1536,0.011722666521867117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,3584,0.007626666790909237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,512,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,1024,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,768,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,768,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,256,0.002159111201763153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,3072,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,512,0.005881777654091517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,512,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,768,0.010330666270520952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,128,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,256,0.005032000028424793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,256,0.005772444523043103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,512,0.00977244476477305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,64,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,512,32,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,128,0.005643555687533484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,256,0.009668444593747457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,128,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,65536,0.024479111035664875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,64,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,128,0.007499555746714274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,32,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,65536,0.04979822370741102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,1536,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,16384,0.008452444440788692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,51200,0.03951644566324022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,51200,0.042867554558648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,65536,0.141792893409729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,12288,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,16384,0.016837333639462788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,16384,0.016513778103722464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,51200,0.11251289314693874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,10240,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,12288,0.013779555757840475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,12288,0.013792888985739814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,16384,0.041801777150895864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,8192,0.00923111124171151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,10240,0.012359999948077731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,10240,0.012618666721714867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,8192,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,8192,0.011039111349317761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,7168,0.009703111317422655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,10240,0.029170665476057265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,51200,0.020344000723626878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,6144,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,7168,0.01034044474363327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,65536,0.05400088760587904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,8192,0.02534488836924235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,7168,0.010060444474220276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,5120,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,6144,0.010033778018421596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,6144,0.009334222310119206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,7168,0.02302933401531643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,4096,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,5120,0.00869066682126787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,5120,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,3584,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,4096,0.00833511104186376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,5120,0.019239111079110038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,4096,0.007703999678293864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,3072,0.005314666777849197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,3584,0.008047111332416534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,3584,0.007637333538797166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,4096,0.01682133310370975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,12288,0.032998220788107976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,2560,0.004872888740566042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,3072,0.0070266665683852295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,3072,0.007698666718271043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,3584,0.015862221519152325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,2048,0.0042408886882993905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,2560,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,2560,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,1536,0.003571555432346132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,2048,0.007096889118353526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,2560,0.013785777820481194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,2048,0.006344888773229387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,1024,0.0030666666312350165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,1536,0.006653333289755716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,1536,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,2048,0.012784889174832238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,1024,0.006520000182920032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,1536,0.011400889191362592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,1024,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,512,0.002463999928699599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,6144,0.020640000700950623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,768,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,512,0.005714666512277391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,512,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,256,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,768,0.010713777608341642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,512,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,128,0.001968888979819086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,256,0.006132444573773279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,256,0.005125333451562458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,3072,0.01514844430817498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,128,0.005653333332803514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,256,0.009004444711738164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,128,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,32,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,65536,0.014675555957688225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,128,0.007699555820888943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,256,768,0.002805333377586471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,64,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,51200,0.012366222010718452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,32,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,65536,0.05168177684148153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,65536,0.04927644464704725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,16384,0.006046222315894232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,768,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,51200,0.03941244549221463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,65536,0.14140533076392278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,51200,0.041533334387673274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,12288,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,16384,0.016520889268981088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,16384,0.016584888100624084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,51200,0.11179644531673855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,10240,0.005984888722499211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,12288,0.014107555150985718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,12288,0.013809777796268463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,16384,0.04176888863245646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,10240,0.012716444830099741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,10240,0.012270222107569376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,12288,0.03317600157525804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,7168,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,8192,0.010599111517270407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,7168,0.009671111073758867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,8192,0.025403555896547105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,6144,0.006074666563007567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,7168,0.010372444159454769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,1024,0.010675555302037133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,7168,0.023016888234350417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,5120,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,6144,0.009670221971140968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,6144,0.009304000271691216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,5120,0.009029332962301042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,6144,0.020608888732062448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,5120,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,4096,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,8192,0.005631999837027655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,3584,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,4096,0.008283555507659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,5120,0.018882667024930317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,4096,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,3072,0.004862222406599256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,8192,0.011062222222487131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,3584,0.007300444775157505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,3584,0.007928888830873702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,4096,0.01685333251953125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,2560,0.005208000126812193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,3072,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,2048,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,3584,0.016329778565300834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,2560,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,2560,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,1536,0.003528000166018804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,2048,0.0070062221752272705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,2560,0.01347733371787601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,1536,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,1536,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,1024,0.002992000016901228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,2048,0.013175110850069257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,1024,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,1536,0.012072888513406118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,768,0.0028408887899584244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,1024,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,10240,0.029476443926493328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,1024,0.011047110789351992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,768,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,512,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,768,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,256,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,512,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,768,0.010728889041476779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,3072,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,128,0.0018586667461527719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,512,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,256,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,3072,0.014811555544535318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,512,0.009648888475365108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,2048,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,128,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,32,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,65536,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,128,0.0070222218831380205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,64,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,51200,0.011756444142924415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,16384,0.007151111132568783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,32,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,12288,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,65536,0.050738665792677135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,10240,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,51200,0.04071822100215488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,8192,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,16384,0.016536888149049547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,7168,0.005411555369695027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,12288,0.013807111316257052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,6144,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,10240,0.012102222277058495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,5120,0.004948444250557157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,8192,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,4096,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,7168,0.010094222095277574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,3584,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,6144,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,256,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,3072,0.004606222112973531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,5120,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,2560,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,128,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,4096,0.007605333295133378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,2048,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,3584,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,1536,0.0034977777136696708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,1024,0.002874666617976295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,3072,0.007270221908887227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,768,0.002593777866827117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,2560,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,512,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,2048,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,256,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,1024,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,128,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,768,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,64,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,256,0.004866666677925321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,64,32,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,128,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,65536,0.009648000200589498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,32,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,51200,0.008632889224423302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,65536,0.05065155691570706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,16384,0.0064106666379504735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,256,0.009687110781669617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,12288,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,51200,0.040700445572535195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,10240,0.0057857777509424425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,16384,0.01651022169325087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,1536,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,8192,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,12288,0.013801777528391944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,6144,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,10240,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,512,0.004969777746333016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,5120,0.004862222406599256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,4096,0.00452533322903845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,8192,0.0107360002067354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,3584,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,6144,0.009378666679064432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,3072,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,64,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,5120,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,2560,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,2048,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,3584,0.007344000041484833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,128,64,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,3072,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,2560,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,2048,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,768,0.002607999990383784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,1536,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,512,0.002465777720014254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,1024,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,256,0.0021848889688650766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,768,0.0053280000057485365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,7168,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,512,0.005078222188684675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,64,0.001783999924858411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,256,0.004964444372389051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,32,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,7168,0.009875555833180746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,128,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,64,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,32,0.0047146665553251905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,1536,0.0035457776652442086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,1024,0.002842666581273079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,16384,0.33713600370619035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,16384,0.26881244447496205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,12288,0.2579102251264784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,16384,0.13492266337076822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,51200,1.0434150695800781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,10240,0.21810666720072427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,12288,0.19929333527882895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,4096,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,8192,0.17658666769663492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,51200,0.40578754742940265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,12288,0.10089511341518825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,10240,0.1657680008146498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,10240,0.08544355630874634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,8192,0.13694577746921116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,6144,0.13619644112057155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,7168,0.15987022717793783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,8192,0.06925333208507962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,7168,0.12303200032975938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,7168,0.06194755766126844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,160,32,128,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,4096,0.09295111232333714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,6144,0.10249155759811401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,5120,0.045801778634389244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,5120,0.0897795557975769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,3584,0.08123555448320177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,4096,0.07295111152860853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,4096,0.03940088881386651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,3072,0.07089689042833117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,2560,0.0598506662580702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,3072,0.030397332376903955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,3584,0.06395288970735338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,3072,0.054899556769265064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,2048,0.04843022094832527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,2560,0.027330666780471802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,2560,0.047352890173594155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,1536,0.03820711043145921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,2048,0.04014488723542955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,5120,0.11513688829210068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,1024,0.028558221128251817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,51200,0.8399999936421713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,1536,0.018575999471876357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,6144,0.05388622151480781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,1536,0.03198044498761495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,768,0.021280889709790547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,512,0.0147724449634552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,1024,0.015328889091809591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,768,0.014162666267818876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,256,0.009128000173303816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,1024,0.024792000651359558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,768,0.030613332986831665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,3584,0.03506844573550754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,128,0.007891555627187094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,256,0.011761777930789523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,64,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,128,0.011603555745548673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,32,0.005793777604897817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,128,0.014601776997248331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,64,0.014656888114081489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,32,0.015303999185562134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,2048,0.023619555764728125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,512,0.013072888884279462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,512,0.02177511155605316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,256,0.015613334046469795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,65536,0.8381910853915744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,51200,0.8882151179843478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,51200,0.6618800163269043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,65536,1.1311261918809679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,16384,0.28639290067884654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,12288,0.21886755360497367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,10240,0.1965573363833957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,16384,0.21278933684031168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,12288,0.08048711220423381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,16384,0.10777066813574897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,12288,0.1580853329764472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,65536,0.4042346742418077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,8192,0.16047199567159018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,10240,0.06865155696868896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,7168,0.1409955554538303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,51200,0.320471101337009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,8192,0.055973331133524575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,7168,0.09691288736131455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,10240,0.13373511367373997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,7168,0.049547556373808116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,5120,0.1024257805612352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,4096,0.08295644654168023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,6144,0.08291466368569268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,6144,0.043552001317342125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,5120,0.03746666510899862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,3584,0.07355822457207574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,5120,0.0707813302675883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,3072,0.06450044446521334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,4096,0.03238933285077413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,3584,0.029304001066419814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,2560,0.05407822132110596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,3584,0.05112711257404751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,4096,0.057087997595469155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,3072,0.044974221123589404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,3072,0.02603911028967963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,1536,0.03413955701722039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,2560,0.039015110996034406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,2048,0.019695111446910434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,2048,0.032885332902272545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,1024,0.024724443753560383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,1536,0.016349333855840895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,1536,0.026874666412671406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,768,0.018538667096032035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,1024,0.013849777479966482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,8192,0.11374310652414958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,6144,0.1209884484608968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,512,0.013243555194801755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,1024,0.020629333125220407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,256,0.008788444101810455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,768,0.012774222426944308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,128,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,256,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,512,0.011424000064531961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,512,0.01846577723821004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,64,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,32,0.0052266667286554975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,128,0.010441777606805166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,768,0.025214221742418077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,128,0.012544000314341651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,64,0.012465777496496836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,32,0.01312622262371911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,2560,0.023390221926901076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,65536,0.3391697671678331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,51200,0.2655271159278022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,65536,0.2982151243421766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,16384,0.0867839985423618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,65536,0.13841422398885092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,12288,0.06806221935484144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,51200,0.2309795485602485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,51200,0.10940800110499065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,16384,0.04012088974316915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,16384,0.07820355229907565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,10240,0.057322667704688184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,12288,0.03161333335770501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,8192,0.04820000131924947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,12288,0.05923910935719808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,256,0.013208889298968844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,10240,0.027411555250485737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,7168,0.04122577773200141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,10240,0.050642665889528066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,8192,0.023212444451120164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,6144,0.035791112316979304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,8192,0.04296444522009956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,7168,0.02199911077817281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,5120,0.030583110120561387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,7168,0.03740888833999634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,2048,0.043967998690075345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,4096,0.025961778230137292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,6144,0.018904000520706177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,6144,0.033217777808507286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,5120,0.016599110431141324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,3584,0.022496889034907024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,3072,0.01998933321899838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,5120,0.0288426677385966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,4096,0.014879110786649915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,2560,0.016521778371598985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,3584,0.01367911116944419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,4096,0.024107555548350017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,2048,0.01349511080318027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,3584,0.021771555145581562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,3072,0.019187556372748483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,2560,0.011931555966536203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,1536,0.010988444089889526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,3072,0.012815111213260226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,2560,0.016996444927321542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,1024,0.008541333178679148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,768,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,1536,0.009325332939624786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,2048,0.014677332507239448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,1024,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,1536,0.012696888711717395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,512,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,256,0.0037884443170494507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,768,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,1024,0.010595555934641095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,128,0.0031893334041039148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,768,0.011427555647161273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,512,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,64,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,256,0.006807111203670502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,512,0.009126221968068017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,32,0.0028782222006056043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,256,0.0073626670572492815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,128,0.006799110935793982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,128,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,64,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,32,0.007834666305118138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,65536,0.2754088772667779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,51200,0.21373955408732095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,16384,0.07068533367580838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,65536,0.21111999617682564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,65536,0.1256151066886054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,12288,0.0554257763756646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,51200,0.16669066747029623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,51200,0.10007199976179336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,2048,0.010777778095669217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,16384,0.03705866800414191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,10240,0.04560533497068617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,16384,0.057019558217790395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,8192,0.03974755605061849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,12288,0.029455112086402044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,7168,0.03388355506791009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,6144,0.029263999727037217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,10240,0.025432889660199482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,8192,0.02188088827662998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,10240,0.03745955559942458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,8192,0.0322364436255561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,5120,0.02525955604182349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,7168,0.028284443749321833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,6144,0.018221333622932434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,5120,0.01607288916905721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,3584,0.018579555882347953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,5120,0.02163377735349867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,4096,0.014461333552996317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,3072,0.016180444094869826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,4096,0.018428444862365723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,3584,0.013740444348918067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,2560,0.013711999687883588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,3584,0.016696888539526198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,3072,0.012421333127551608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,2048,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,3072,0.015043555034531487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,2560,0.011417778001891242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,1536,0.009684444301658207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,2560,0.01368799971209632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,2048,0.009744000103738572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,2048,0.01241422196229299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,1024,0.00792266676823298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,1536,0.010581333604123859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,7168,0.019768888751665752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,768,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,4096,0.021665778425004747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,1024,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,1024,0.008615111311276754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,512,0.005041777673694822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,6144,0.02509422269132402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,256,0.00387999994887246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,768,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,512,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,768,0.009686222506894005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,128,0.0030008889734745026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,256,0.00683111117945777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,64,0.0028320000403457214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,256,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,32,0.0028844444702068963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,128,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,64,0.00665511108107037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,12288,0.043615112702051796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,32,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,65536,0.26648977067735463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,51200,0.2082657814025879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,65536,0.20047999752892387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,65536,0.1223013268576728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,16384,0.06928444570965238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,51200,0.15597066614362928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,512,0.00796977761718962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,51200,0.0975724458694458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,16384,0.03561955690383911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,10240,0.04530310961935255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,128,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,16384,0.05428444345792135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,12288,0.02829066581196255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,12288,0.041682667202419706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,8192,0.037809779246648155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,7168,0.033165332343843244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,10240,0.024448888169394598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,8192,0.02110222147570716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,10240,0.03603555427657233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,6144,0.028871999846564397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,8192,0.030591110388437908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,7168,0.019137778215938143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,7168,0.02684533264901903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,1536,0.008270222279760573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,4096,0.020786666207843356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,6144,0.01741066740618812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,3584,0.018248889181349013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,5120,0.01590222285853492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,4096,0.014119111829333834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,5120,0.02034222251839108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,3072,0.016146666473812528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,4096,0.017675555414623685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,3584,0.013096888860066732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,2560,0.013663999736309052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,3584,0.016160888804329764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,3072,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,3072,0.014774221513006421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,12288,0.05417422122425503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,2560,0.013281777501106262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,1536,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,2048,0.008986666798591614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,2048,0.012034667034943899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,1024,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,1536,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,768,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,1536,0.010283555421564314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,1024,0.007528000407748752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,512,0.004978666702906291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,1024,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,768,0.007072000039948358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,5120,0.02476444509294298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,256,0.003842666745185852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,768,0.009043555292818282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,512,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,128,0.0030951110853089225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,512,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,64,0.002983111060327954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,128,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,256,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,32,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,128,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,2048,0.011716444459226398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,64,0.006311999840868845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,65536,0.18359643883175322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,32,0.006679111056857639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,51200,0.14363111390007868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,65536,0.1586666636996799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,16384,0.051945779058668345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,65536,0.07389066616694133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,51200,0.05990755558013916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,51200,0.12310043970743816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,12288,0.04021688964631822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,16384,0.023694222172101338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,10240,0.033238222201665245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,16384,0.043699555926852755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,8192,0.028350220786200628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,10240,0.016855110724767048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,256,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,10240,0.029127111037572224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,8192,0.014875556031862894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,7168,0.02459200057718489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,8192,0.024123556084103052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,6144,0.021549334128697712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,7168,0.013705777625242868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,2560,0.010492444038391113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,7168,0.02183377742767334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,5120,0.018538667096032035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,6144,0.019446222318543326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,4096,0.015691555208630033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,5120,0.011628444823953839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,5120,0.017098666893111337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,3584,0.013463999662134381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,4096,0.010681777364677854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,4096,0.015054222610261707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,3072,0.012206222448084088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,12288,0.01891999940077464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,3584,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,3584,0.013769778112570444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,2560,0.010423111418883005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,3072,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,6144,0.023631999890009563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,2048,0.008981333010726506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,3072,0.012735111018021902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,1536,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,2560,0.011612444288200803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,2048,0.01036088913679123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,1024,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,1536,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,1536,0.008969777988062965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,768,0.00452622233165635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,6144,0.012740444805887012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,1024,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,512,0.0038737778862317405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,1024,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,768,0.00655733338660664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,256,0.0029191111938820947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,768,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,512,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,128,0.00253866674999396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,512,0.006828444285525217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,256,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,64,0.0024702221983008915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,256,0.006197333335876465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,12288,0.0337119996547699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,128,0.006135111053784688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,32,0.0026195556339290407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,128,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,64,0.006187555690606435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,2560,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,32,0.0063075555695427795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,2048,0.0073626670572492815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,51200,0.1307706700430976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,65536,0.1535608900917901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,16384,0.04900177650981479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,65536,0.07119288709428576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,12288,0.03863644599914551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,51200,0.057645334137810596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,16384,0.022480888499153986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,10240,0.032264000839657254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,16384,0.04241333405176798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,8192,0.02751200066672431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,12288,0.032891554964913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,10240,0.016488000750541687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,7168,0.024059555596775476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,10240,0.028303109937243994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,8192,0.01479111115137736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,6144,0.020671111014154222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,8192,0.0236471113231447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,7168,0.013420444395807056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,5120,0.017656000124083627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,7168,0.021055110626750525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,6144,0.01237777786122428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,4096,0.01477599971824222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,5120,0.01129422254032559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,3584,0.013068444199032254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,5120,0.01684888866212633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,4096,0.010055999788973067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,65536,0.16594489415486655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,3072,0.011959111524952782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,4096,0.014710222681363424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,3584,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,2560,0.01013777818944719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,3072,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,3584,0.0136817776494556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,2048,0.008991999758614434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,2560,0.007678222325113084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,51200,0.12088444497850205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,1536,0.007276444799370236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,2048,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,2560,0.011412444214026133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,1024,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,2048,0.009894222021102905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,768,0.004530666602982415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,1536,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,1536,0.008296888735559251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,512,0.0038480001191298165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,1024,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,768,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,768,0.008249777886602614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,256,0.0029057777590221832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,512,0.006106666806671355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,512,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,6144,0.019069333871205647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,256,0.005882666756709416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,256,0.006070222291681502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,64,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,128,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,32,0.0024986666523747975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,128,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,3072,0.012601777911186218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,64,0.005972444597217772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,65536,0.1486862235599094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,32,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,12288,0.018615111708641052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,51200,0.1188453303443061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,16384,0.04565244581964281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,1024,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,65536,0.06881244315041436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,65536,0.11504977279239231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,12288,0.03373510970009698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,51200,0.0903351108233134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,51200,0.056177775065104164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,10240,0.026935999592145283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,16384,0.021623111433453027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,8192,0.023226666781637404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,12288,0.017839110559887357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,12288,0.025425778494940862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,128,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,7168,0.01996977792845832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,8192,0.014096889230940076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,10240,0.02197155521975623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,6144,0.017855111095640395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,8192,0.01869777838389079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,7168,0.01313333296113544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,5120,0.014922666880819531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,7168,0.016893333858913846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,4096,0.012840888566441007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,6144,0.01205244412024816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,3584,0.011363555159833698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,6144,0.01551822159025404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,5120,0.010878221856223213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,3072,0.01036177741156684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,4096,0.009577777650621202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,5120,0.01385600037044949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,2560,0.009232000344329411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,4096,0.012449777788586087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,3072,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,3584,0.011709333293967776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,2048,0.007682667010360294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,3072,0.01072177787621816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,2560,0.007612444460391998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,1536,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,2560,0.009331555830107795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,2048,0.007245333658324347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,1024,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,1536,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,2048,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,768,0.004223111189073986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,1536,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,16384,0.03279022375742594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,512,0.003448888866437806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,10240,0.015995555453830294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,768,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,1024,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,768,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,128,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,256,0.0058746664888328975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,512,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,64,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,256,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,32,0.002552888873550627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,128,0.0058808889653947614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,128,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,3584,0.0087013335691558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,65536,0.11751200093163384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,64,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,32,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,51200,0.09509955512152778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,16384,0.03696889016363356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,65536,0.06671644581688775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,51200,0.05355199840333727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,51200,0.08914844195048015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,12288,0.03253777821858724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,16384,0.020966221888860066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,1024,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,10240,0.027622222900390625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,16384,0.0315324432320065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,256,0.0027955555253558685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,12288,0.024615110622511968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,8192,0.022699554761250813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,10240,0.01550222271018558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,7168,0.019720888800091214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,10240,0.02126577827665541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,8192,0.013654222091039022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,6144,0.0177182224061754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,8192,0.018439110782411363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,7168,0.012757333616415659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,5120,0.014583110809326172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,7168,0.016607999801635742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,6144,0.011701333026091257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,4096,0.012530666258600024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,6144,0.015187554889255099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,5120,0.010122666756312052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,5120,0.013618666264745923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,3584,0.011125333607196808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,4096,0.009139555195967356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,65536,0.11393333805931939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,3072,0.00998222248421775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,3584,0.008359111017651027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,2560,0.008727110922336578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,3584,0.011112000379297467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,3072,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,2048,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,3072,0.010337777435779572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,2560,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,2560,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,12288,0.017335999343130324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,1536,0.006224000205596288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,1024,0.004842666702138053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,2048,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,512,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,1536,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,2048,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,1536,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,512,0.0031911111954185697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,1024,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,768,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,256,0.0028488888508743713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,768,0.006903111106819577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,512,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,128,0.0025066667132907445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,512,0.006162666612201267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,256,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,256,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,64,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,32,0.002198222196764416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,4096,0.012099555797047086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,128,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,65536,0.09285422166188557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,64,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,32,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,51200,0.0765626695421007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,16384,0.033011555671691895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,65536,0.10644355747434829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,51200,0.05139466789033678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,768,0.003867555409669876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,12288,0.02572444412443373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,16384,0.02019377715057797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,1024,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,10240,0.020991111795107525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,16384,0.02988977895842658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,12288,0.016812443733215332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,8192,0.017655111021465726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,10240,0.015156444576051501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,12288,0.023648000425762598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,7168,0.015405333704418607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,10240,0.020599999361568026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,8192,0.013170666992664337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,6144,0.013354666531085968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,8192,0.017782222893502977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,7168,0.012081777883900536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,5120,0.011528000235557556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,128,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,7168,0.016146666473812528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,6144,0.010672888822025724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,4096,0.010311110979980892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,3584,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,6144,0.014803555276658801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,5120,0.00923644420173433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,4096,0.008657777474986183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,5120,0.013442666994200813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,3072,0.0075839997993575195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,2560,0.006789333290523953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,4096,0.011700444751315646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,3584,0.008204444415039485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,2048,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,3584,0.01053422192732493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,3072,0.007752888732486301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,65536,0.06395555867089166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,1536,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,2560,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,3072,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,1024,0.0038355555799272326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,51200,0.08322578006320529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,2560,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,768,0.0032044444233179092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,2048,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,1024,0.0064479998416370815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,1536,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,768,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,1024,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,256,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,768,0.006351111249791251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,512,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,128,0.0025066667132907445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,512,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,256,0.005724444571468566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,64,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,256,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,32,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,128,0.0053093334039052325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,64,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,65536,0.08506933185789321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,32,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,51200,0.06981066862742107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,65536,0.06288799974653456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,65536,0.10360088613298203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,2048,0.007330666813585494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,16384,0.03161599900987413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,1536,0.006842666616042455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,51200,0.05024977856212192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,51200,0.08138222164577909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,12288,0.02530844509601593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,16384,0.01993333299954732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,10240,0.022089777721299067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,16384,0.02952622373898824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,12288,0.016540444559521146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,8192,0.016735111673672993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,10240,0.014915555715560913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,12288,0.023192000057962205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,7168,0.014673777752452426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,10240,0.020297777321603563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,128,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,8192,0.012888000243239932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,6144,0.012664000193277994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,7168,0.011552888486120435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,5120,0.011124444504578909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,6144,0.010677333507272931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,7168,0.015871110889646742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,4096,0.009909333454238044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,6144,0.014474666780895658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,5120,0.009335111412737105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,3584,0.008628444539176093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,5120,0.013096888860066732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,3072,0.007861333588759104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,4096,0.011250666446155973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,3584,0.008220444122950235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,2560,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,3584,0.010044444766309526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,3072,0.00793511089351442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,2048,0.005647999958859549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,3072,0.009388444324334463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,2560,0.007600000335110559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,1536,0.004857777721352047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,2560,0.008422222402360704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,1024,0.0036524443162812125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,1536,0.006693333387374878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,2048,0.007701333198282454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,768,0.0033253334048721525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,1536,0.007096889118353526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,1024,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,512,0.0028888889484935333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,1024,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,768,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,256,0.0024844445288181305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,8192,0.017447999782032438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,512,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,128,0.0022542222092549005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,512,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,256,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,4096,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,64,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,128,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,32,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,128,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,65536,0.07448799742592706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,64,0.005396444350481033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,32,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,512,0.002887111157178879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,51200,0.06025511026382446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,65536,0.06208799944983589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,2048,0.007220444579919179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,16384,0.024559110403060913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,12288,0.020263999700546265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,51200,0.049264887968699135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,16384,0.019561777512232464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,51200,0.051333334710862905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,10240,0.017490666773584154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,16384,0.019895111521085102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,12288,0.01646844380431705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,8192,0.0141004439857271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,12288,0.0162791113058726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,768,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,7168,0.012488000094890594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,10240,0.01440622243616316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,8192,0.012373333175977072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,6144,0.011119999819331698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,256,0.005642666584915585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,7168,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,5120,0.010016889207892949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,7168,0.011685333318180509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,6144,0.010016889207892949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,4096,0.008604444563388824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,5120,0.009303111169073317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,3584,0.007848000360859765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,5120,0.010055999788973067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,4096,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,3072,0.00721955547730128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,4096,0.00849511143234041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,3584,0.008064000142945183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,3584,0.008012444608741337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,2560,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,3072,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,3072,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,2048,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,2560,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,2560,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,1536,0.004371555729044808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,2048,0.007343110938866933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,10240,0.014689778288205465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,1024,0.003543111185232798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,1536,0.00674044465025266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,8192,0.012794666820102267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,768,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,1024,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,1024,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,512,0.0028773333049482773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,768,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,768,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,256,0.0024702221983008915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,512,0.005712000032265981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,65536,0.0654746691385905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,128,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,256,0.005649777750174205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,256,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,64,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,128,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,32,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,128,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,65536,0.06488711304134794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,64,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,2048,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,51200,0.05271377828386095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,32,0.005231110999981563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,1536,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,16384,0.023039110832744177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,65536,0.06021510892444187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,65536,0.06320266591178046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,12288,0.018923555811246235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,51200,0.048340443107816905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,51200,0.04915644394026863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,16384,0.01942488882276747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,10240,0.016594666573736403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,8192,0.01352533366945055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,12288,0.015827554795477126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,12288,0.015832889411184523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,10240,0.013815999858909182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,7168,0.01178311142656538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,512,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,8192,0.011679111255539788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,6144,0.010811555716726514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,8192,0.01238044434123569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,7168,0.010751999914646149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,5120,0.009663999908500248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,7168,0.011387555963463254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,6144,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,4096,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,6144,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,5120,0.009085333181752099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,3584,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,4096,0.00851466672288047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,5120,0.009572444690598382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,4096,0.00812266684240765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,3584,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,2560,0.006199999815887875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,3072,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,3072,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,2048,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,6144,0.010849778023031024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,2560,0.0075902218619982404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,1536,0.0041751112374994485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,2560,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,16384,0.01906755566596985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,2048,0.007289778027269576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,1024,0.003620444486538569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,2048,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,768,0.0031582222630580268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,1536,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,10240,0.013836444252067141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,1536,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,512,0.002846222163902389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,768,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,1024,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,256,0.002500444443689452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,512,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,768,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,128,0.0021582223061058256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,512,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,256,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,64,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,32,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,128,0.005631999837027655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,256,0.005203555441564984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,3584,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,128,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,65536,0.053136888477537364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,51200,0.04266044497489929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,64,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,32,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,65536,0.059079110622406006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,16384,0.018560889694425795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,51200,0.047276443905300565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,16384,0.01890222231547038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,12288,0.014732443624072604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,16384,0.018608000543382432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,10240,0.012823999755912356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,12288,0.015322667029168872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,1024,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,12288,0.015392889579137167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,8192,0.012549333274364471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,10240,0.013041777743233575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,7168,0.011338666909270816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,10240,0.01368799971209632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,8192,0.011373333632946014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,6144,0.010616000327799054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,8192,0.011873777541849347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,7168,0.010671111444632212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,3072,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,7168,0.011211555865075854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,6144,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,6144,0.010020444790522257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,5120,0.009115555220180089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,5120,0.008945778012275696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,65536,0.060862223307291664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,3584,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,4096,0.008364444805516137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,51200,0.047154668304655284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,3072,0.006671999891599019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,4096,0.00812888890504837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,2560,0.00563733321097162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,3072,0.007631111476156447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,3584,0.007851555943489075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,2048,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,3072,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,2560,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,2560,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,1536,0.004549333204825719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,2048,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,1024,0.0035422220826148987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,2048,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,1536,0.006664888726340399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,768,0.003162666741344664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,1536,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,1024,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,5120,0.009387555221716562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,512,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,1024,0.005665777872006099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,768,0.006047111004590988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,256,0.00245066670080026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,768,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,512,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,128,0.0021893332401911416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,256,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,64,0.002120888895458645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,512,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,32,0.0021333334346612296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,128,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,256,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,128,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,3584,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,51200,0.03455555438995361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,32,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,65536,0.057687997817993164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,65536,0.057782219515906445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,16384,0.015544000599119397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,51200,0.046386665768093534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,51200,0.045528001255459256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,12288,0.013280889226330651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,16384,0.01829155617290073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,16384,0.018262222409248352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,10240,0.01200355589389801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,12288,0.014126222994592456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,8192,0.011019555230935415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,12288,0.015101333459218344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,10240,0.012773333324326409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,7168,0.011952888634469775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,10240,0.013432888521088494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,4096,0.008295111358165741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,8192,0.011367110742463006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,6144,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,5120,0.008980444735950893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,8192,0.011039999624093374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,7168,0.010435555544164447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,6144,0.009995555712117089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,4096,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,6144,0.009532444179058075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,65536,0.04339822133382162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,3584,0.00703288863102595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,5120,0.008797333472304875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,4096,0.008330666356616551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,3072,0.005974222388532426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,64,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,2560,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,4096,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,3584,0.008024888734022776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,2048,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,3072,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,3584,0.007637333538797166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,2560,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,3072,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,1536,0.0041306668685542215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,2048,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,2560,0.006994666324721442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,1024,0.0031848889258172777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,2048,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,1536,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,768,0.002924444360865487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,1024,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,1536,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,512,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,1024,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,768,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,256,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,768,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,512,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,128,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,512,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,256,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,64,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,256,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,32,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,128,0.005797333187527127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,7168,0.010568000376224518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,65536,0.03182844320933024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,5120,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,128,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,51200,0.02587555514441596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,16384,0.013007111019558378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,64,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,65536,0.056919998592800565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,32,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,12288,0.011358222199810876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,51200,0.045893334680133395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,10240,0.01033866653839747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,16384,0.016859556237856548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,51200,0.04348355531692505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,8192,0.009454222189055549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,16384,0.017633777525689866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,12288,0.013826666606797112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,7168,0.009002666506502364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,8192,0.011124444504578909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,8192,0.010739555789364709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,6144,0.008260444634490544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,7168,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,7168,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,5120,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,6144,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,6144,0.009401777552233802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,4096,0.006907555378145642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,5120,0.009000889129108852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,5120,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,3584,0.006310222049554189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,4096,0.008340444829728868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,4096,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,65536,0.05503733290566338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,3072,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,3584,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,3584,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,12288,0.01387644476360745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,3072,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,2048,0.004515555583768421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,10240,0.012784000072214337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,3072,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,2560,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,10240,0.01239822225438224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,2560,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,2048,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,2048,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,1536,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,768,0.002871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,1024,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,1536,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,512,0.002546666603949335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,1024,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,256,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,768,0.006138666636413998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,512,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,128,0.0021253333737452826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,512,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,256,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,64,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,128,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,32,0.0018568889548381169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,128,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,65536,0.02642311155796051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,64,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,32,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,51200,0.02212000058756934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,1536,0.0038613333470291565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,65536,0.056530667675866016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,1024,0.0031742221779293488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,16384,0.011192888849311404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,51200,0.04590577880541483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,12288,0.009705777797434065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,51200,0.04310755597220527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,16384,0.017276444368892245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,10240,0.008997333546479544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,12288,0.013762666947311826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,768,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,12288,0.013759111364682516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,8192,0.008184889124499427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,10240,0.0127688886390792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,10240,0.012096888489193387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,7168,0.00959288908375634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,8192,0.011115555961926779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,8192,0.010682666467295753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,6144,0.008636444807052612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,7168,0.01053955571519004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,7168,0.010030222435792288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,5120,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,6144,0.009708444277445475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,65536,0.05384177631802029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,4096,0.007104000283612146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,5120,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,5120,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,16384,0.01663466625743442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,4096,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,4096,0.007930667036109501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,3072,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,3584,0.008030222521887885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,3584,0.007532444265153673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,2560,0.005161777966552311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,3072,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,3072,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,2048,0.004515555583768421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,2560,0.007384888827800751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,2560,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,2560,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,1536,0.0038702223036024305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,2048,0.0069857777820693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,2048,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,1536,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,1024,0.0031893334041039148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,768,0.002883555574549569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,1536,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,6144,0.009307555854320526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,1024,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,512,0.002481777841846148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,768,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,1024,0.005736888696750005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,256,0.0022053333620230355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,3584,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,768,0.005442666510740916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,512,0.005884444548024072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,128,0.001971555459830496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,64,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,512,0.0053031109273433685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,768,32,0.001963555605875121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,128,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,256,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,65536,0.022133333815468684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,128,0.004956444518433677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,51200,0.017804443836212158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,64,0.005011555635266834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,16384,0.011710222396585675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,65536,0.05621244510014852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,32,0.005038222091065513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,12288,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,51200,0.042790220843421094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,65536,0.05281155639224582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,51200,0.042486220598220825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,16384,0.016501333978441026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,8192,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,16384,0.016507556041081745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,10240,0.012467555701732635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,12288,0.013754666679435305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,7168,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,10240,0.012103999654452005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,8192,0.011066666907734342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,6144,0.006679111056857639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,8192,0.010649777948856354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,5120,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,6144,0.009682666924264696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,7168,0.01000533335738712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,4096,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,6144,0.009325332939624786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,5120,0.008992888861232335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,256,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,3584,0.006214222146405115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,5120,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,3072,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,4096,0.007879111501905654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,3584,0.00795022232664956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,2560,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,3584,0.007575111256705389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,3072,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,3072,0.007300444775157505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,2048,0.004230222354332606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,2560,0.007335111498832703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,10240,0.009002666506502364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,1536,0.0037697777152061462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,2560,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,12288,0.014115555418862237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,1024,0.003136000078585413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,2048,0.006483555667930179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,1536,0.006647111227114995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,768,0.0028391112056043413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,1024,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,1536,0.006014222072230444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,512,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,1024,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,7168,0.01036533299419615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,256,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,768,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,768,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,128,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,512,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,64,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,256,0.00509333320789867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,512,32,0.001871111078394784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,4096,0.008347555167145198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,128,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,65536,0.018664000762833487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,64,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,65536,0.049527999427583486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,32,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,65536,0.051783111360338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,16384,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,2048,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,51200,0.041800889703962535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,12288,0.011376000112957425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,16384,0.01648977730009291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,16384,0.016348444753222995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,10240,0.01036800030204985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,12288,0.013769778112570444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,12288,0.013574221895800697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,512,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,8192,0.008961777720186446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,10240,0.012505778008037143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,10240,0.012069332930776807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,7168,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,8192,0.011086222198274402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,6144,0.007720000214046902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,128,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,8192,0.010650667051474253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,5120,0.00701333334048589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,7168,0.009952888720565373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,6144,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,51200,0.015824000040690105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,4096,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,6144,0.009314667019579146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,5120,0.008960000342792934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,3584,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,5120,0.008664888640244802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,4096,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,3072,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,3584,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,4096,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,2560,0.004863111095296012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,3584,0.007330666813585494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,2048,0.004212444441186057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,3072,0.00721155520942476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,2560,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,1536,0.003539555602603488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,2048,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,1024,0.003206222214632564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,2048,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,1536,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,768,0.0028044444819291434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,1024,0.006224000205596288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,7168,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,1024,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,512,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,768,0.006102222121424145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,256,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,768,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,512,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,128,0.0018666666001081467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,512,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,256,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,64,0.0017893332988023758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,51200,0.039493332306543984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,256,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,256,32,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,128,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,65536,0.014813333749771118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,128,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,3072,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,51200,0.012723555167516073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,64,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,2560,0.007280888656775157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,32,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,16384,0.008739555875460306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,65536,0.048913776874542236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,12288,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,65536,0.04935288760397169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,10240,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,16384,0.016521778371598985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,51200,0.03990666733847724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,12288,0.013802666631009845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,8192,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,16384,0.016227554943826463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,12288,0.013108444710572561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,7168,0.006599111275540457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,10240,0.01242844429281023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,1536,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,6144,0.007252444823582967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,10240,0.011752000285519494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,7168,0.010401777923107147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,8192,0.010061333576838175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,5120,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,6144,0.009630222287442949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,7168,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,4096,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,6144,0.009024889104896123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,5120,0.008956444760163626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,3584,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,4096,0.008042666647169325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,5120,0.00832266691658232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,3072,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,4096,0.007361777954631382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,2560,0.004844444493452708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,3584,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,3072,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,2048,0.004212444441186057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,3072,0.006873777757088344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,2560,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,1536,0.0035075553589397003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,51200,0.03934489025010003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,2560,0.006506666541099548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,2048,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,2048,0.005993777679072485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,1536,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,768,0.0028311111446883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,1536,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,1024,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,512,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,1024,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,768,0.006134222365087933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,256,0.0021297778520319196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,768,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,512,0.005730666634109285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,128,0.0018622221218215094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,512,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,256,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,64,0.001800000046690305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,256,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,128,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,32,0.0018088887963030075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,3584,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,65536,0.018702222241295707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,51200,0.016153777639071148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,64,0.004601777841647466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,16384,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,32,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,12288,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,65536,0.04894311229387919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,10240,0.006167999986145232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,128,1024,0.0028942223224374983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,51200,0.039705776506000094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,8192,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,7168,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,16384,0.015985776980717976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,6144,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,12288,0.013471999930010902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,5120,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,10240,0.011766222616036734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,8192,0.01036266651418474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,4096,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,3584,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,7168,0.009879111415810054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,3072,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,6144,0.008977777428097194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,2560,0.004863111095296012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,5120,0.008343111309740279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,2048,0.004594666676388847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,4096,0.007782222496138678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,3584,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,1024,0.002893333426780171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,3072,0.006866666591829724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,128,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,2560,0.006360000206364526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,512,0.002475555572244856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,2048,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,256,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,1536,0.005956444475385878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,128,0.0018115554832749898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,1024,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,64,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,768,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,32,0.0018071112119489247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,256,0.004941333499219683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,65536,0.021021333005693223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,64,0.004629333400064045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,51200,0.018181333939234417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,32,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,8192,0.011062222222487131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,16384,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,12288,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,65536,0.04903822143872579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,10240,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,51200,0.03939555419815911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,8192,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,16384,0.01607911123169793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,1536,0.0035031110876136353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,7168,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,12288,0.01348888874053955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,10240,0.011713777979214987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,6144,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,5120,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,64,768,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,4096,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,8192,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,7168,0.009725333087974125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,3584,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,3072,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,6144,0.008995555341243744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,5120,0.008442666795518663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,512,0.005055111315515307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,4096,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,2048,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,3584,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,128,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,3072,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,1024,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,768,0.002604444407754474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,2560,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,512,0.002504000026318762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,2048,0.006005333529578314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,256,0.002036444428894255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,1536,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,128,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,768,0.005530666559934616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,64,0.0017946666727463405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,512,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,32,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,256,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,128,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,64,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,32,0.004975999808973736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,2560,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,129,32,1536,0.0035075553589397003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,16384,0.37349332703484434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,51200,0.624741342332628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,1024,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,16384,0.20535733964708117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,51200,1.153201738993327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,12288,0.27986489401923287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,12288,0.15141956011454263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,16384,0.1357146633995904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,10240,0.2369848887125651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,51200,0.39801687664455837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,10240,0.12801422013176814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,8192,0.199217783080207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,10240,0.08601511187023586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,8192,0.06993689139684041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,7168,0.17034310764736602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,12288,0.22287376721700033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,12288,0.10231200191709731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,16384,0.293250666724311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,10240,0.18924266762203643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,7168,0.09150489171346028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,8192,0.1521786716249254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,6144,0.1455679999457465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,7168,0.06243111027611626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,6144,0.07988799942864312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,5120,0.12325510713789199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,6144,0.05492711067199707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,5120,0.0699475540055169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,4096,0.09902933571073745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,7168,0.13369511233435735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,8192,0.10802221960491604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,3584,0.08783911334143744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,4096,0.040001776483323835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,6144,0.11714933978186713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,5120,0.0990062223540412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,4096,0.08095733324686687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,3072,0.0768702228864034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,3072,0.043396443128585815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,3584,0.07288977834913465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,3072,0.03166310985883077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,3072,0.06588977575302124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,2560,0.06445866823196411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,5120,0.04591644472546048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,2048,0.052114665508270264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,2560,0.03801688883039687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,4096,0.05637422535154554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,2048,0.03289599882231818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,2048,0.023939554889996845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,3584,0.04974755644798279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,3584,0.03485066692034403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,1536,0.04092622134420607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,2560,0.05771911144256592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,1024,0.02941511074701945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,2048,0.04944533440801832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,1536,0.025839999318122864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,768,0.021998221675554912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,1024,0.020046222541067336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,1024,0.01589777734544542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,1536,0.04163377814822727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,1024,0.0340266658200158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,512,0.015249777171346875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,768,0.014160888062583076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,768,0.016199111938476562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,256,0.009782222410043081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,512,0.012873777912722694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,512,0.013682666752073499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,2560,0.028018666638268366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,256,0.012488889197508493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,512,0.025076443950335186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,256,0.011875555747085147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,256,0.0207360006040997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,128,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,1536,0.019321777754359774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,64,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,32,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,128,0.011529777612951068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,128,0.016511110795868766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,64,0.011366222467687396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,51200,0.9298569361368815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,32,0.01477866702609592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,768,0.03049244483311971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,128,0.007255111303594377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,65536,0.638154665629069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,65536,1.1179030736287434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,51200,0.4883315298292372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,16384,0.28204443719651967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,16384,0.16272800498538548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,65536,0.4098009003533258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,51200,0.3195822238922119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,12288,0.21537600623236763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,16384,0.10922844542397393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,51200,0.8773546748691134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,12288,0.12534222337934706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,10240,0.1677875518798828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,12288,0.08240444130367704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,8192,0.1395262214872572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,10240,0.06930310858620538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,10240,0.10209066338009304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,8192,0.08776711093054877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,12288,0.18127911620669893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,7168,0.12032088968488906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,16384,0.2385662131839328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,8192,0.056031107902526855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,6144,0.10352444648742676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,10240,0.15659556123945448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,7168,0.05361688799328274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,7168,0.0750524467892117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,5120,0.08902044428719415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,6144,0.04443466663360596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,7168,0.10915999942355686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,5120,0.054253333144717746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,6144,0.09723733531104194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,5120,0.037463999456829496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,4096,0.07357333103815715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,5120,0.08197955290476482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,4096,0.033025778002209134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,4096,0.044904887676239014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,3584,0.06343022320005628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,4096,0.06934755378299289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,3584,0.029139555162853662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,3584,0.04056355688307021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,3072,0.05548888776037428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,6144,0.06405511167314318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,3072,0.025820444027582806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,3584,0.06253866354624431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,2560,0.04686133397950066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,8192,0.1254026624891493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,2048,0.03844088978237576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,51200,0.739407115512424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,2560,0.02336888843112522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,3072,0.05713333023918999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,1536,0.029652443197038438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,2560,0.03115733464558919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,2048,0.02719733284579383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,2048,0.02047644390000237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,2560,0.04956000049908956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,1024,0.0222382214334276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,1536,0.016759110821617972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,1536,0.021681777305073206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,2048,0.04749155706829495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,768,0.01686311099264357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,1024,0.016539555456903245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,1536,0.0364951127105289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,512,0.011386666860845355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,65536,0.9666133456759982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,1024,0.03046577837732103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,768,0.013769778112570444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,256,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,3072,0.035467555125554405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,512,0.0114720000161065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,512,0.012063999970753988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,128,0.005727111051479976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,256,0.010792888700962067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,512,0.024079110887315538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,128,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,256,0.018921777606010437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,64,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,32,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,128,0.010429333481523724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,64,0.010041777458455827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,32,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,128,0.014494222071435718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,1024,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,768,0.012679999901188744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,65536,0.3249582184685601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,51200,0.24977156851026747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,65536,0.23160711924235025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,768,0.026312889324294195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,16384,0.08673066563076443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,65536,0.1423813369539049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,256,0.01033866653839747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,51200,0.17689243952433267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,12288,0.06541777981652154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,16384,0.040622221099005804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,16384,0.0629075566927592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,12288,0.0322728885544671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,10240,0.055973331133524575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,12288,0.046648889780044556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,12288,0.04794222116470337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,10240,0.0402728882100847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,8192,0.0461404456032647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,10240,0.028135998381508723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,10240,0.04213688770929972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,65536,0.20955467224121094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,51200,0.16570488611857095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,8192,0.034854223330815635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,7168,0.03974399964014689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,8192,0.023885332875781592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,51200,0.11335466967688666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,6144,0.0350471105840471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,8192,0.03640444411171807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,7168,0.03015822172164917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,5120,0.029636445972654555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,16384,0.06169333060582479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,6144,0.02035466664367252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,7168,0.0328693323665195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,4096,0.026347556047969397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,5120,0.017135999268955655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,6144,0.030460443761613634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,3584,0.022123555342356365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,4096,0.015446222490734525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,4096,0.019671999745898776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,5120,0.026868444350030687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,3072,0.019211555520693462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,3584,0.014765333798196582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,4096,0.02386666668785943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,3584,0.017553778158293832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,2560,0.016501333978441026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,3072,0.013199999928474426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,3584,0.021324444148275588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,3072,0.015791111522250705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,2048,0.013384000294738345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,7168,0.022385777698622808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,2560,0.01390044391155243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,2560,0.012095111111799875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,3072,0.019557333654827543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,1536,0.010746666954623329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,2048,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,2560,0.01854933301607768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,1024,0.008360000120268928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,1536,0.010836444795131683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,1536,0.009573332965373993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,5120,0.023250665929582384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,2048,0.017367111312018502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,768,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,1536,0.015107555521859063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,1024,0.008258666429254744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,512,0.005480888817045424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,768,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,768,0.008106666306654612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,1024,0.014067555467287699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,256,0.0038720000949170855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,512,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,512,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,768,0.012545777691735161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,128,0.0031475555151700974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,256,0.006776000062624614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,256,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,512,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,64,0.002874666617976295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,128,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,128,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,32,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,256,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,2048,0.011367110742463006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,64,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,128,0.00832266691658232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,65536,0.24392443233066133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,32,0.007670222057236566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,1024,0.009089777866999308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,51200,0.18675377633836535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,65536,0.17654400401645234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,16384,0.06342489189571805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,65536,0.12616356213887533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,51200,0.13603111108144125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,12288,0.04943644338183933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,16384,0.048927999205059476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,51200,0.10152266422907512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,16384,0.03752177622583177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,10240,0.04201066825124952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,12288,0.02980888883272807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,12288,0.037336889240476824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,16384,0.05266222357749939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,8192,0.03499644332461887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,10240,0.032116442918777466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,6144,0.02683555583159129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,10240,0.02571822206179301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,12288,0.04252622193760342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,51200,0.13778755399915907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,7168,0.030052443345387776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,8192,0.028216001060273912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,65536,0.1731733348634508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,8192,0.021930666433440313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,10240,0.037236445479922824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,6144,0.02638133366902669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,7168,0.019910222954220243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,8192,0.032688889238569475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,7168,0.024401777320437964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,6144,0.018645332919226754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,6144,0.021555556191338435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,4096,0.019645333290100098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,7168,0.029667556285858154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,5120,0.01614488826857673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,6144,0.02730133301681942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,3584,0.01685244507259793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,4096,0.014529777897728814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,4096,0.01611199975013733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,5120,0.02427911096149021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,3072,0.014534221755133735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,3584,0.014136888914638095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,4096,0.021143111917707656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,3072,0.012778666284349231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,3584,0.019714666737450492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,3072,0.013345777988433838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,2560,0.012364444633324942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,3072,0.01831822262869941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,2048,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,2560,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,5120,0.022936888866954382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,2560,0.017200888858901132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,2048,0.010417777631017896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,1536,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,2048,0.011071999867757162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,5120,0.01847644481394026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,1024,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,1536,0.008712000317043727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,2048,0.016120889120631747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,1536,0.009723555710580613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,768,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,1024,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,3584,0.014497778481907315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,1024,0.007848000360859765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,512,0.004287999951177173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,768,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,1024,0.012990222209029727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,512,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,256,0.0032924444725116095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,512,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,2560,0.011483555866612328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,768,0.011404444773991903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,128,0.003156444471743372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,256,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,512,0.011008888483047485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,256,0.006140444427728653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,64,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,128,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,32,0.00318755561278926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,128,0.006647111227114995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,256,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,128,0.008088000118732452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,65536,0.2025360001458062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,64,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,32,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,51200,0.15643910566965738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,65536,0.16882667276594374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,1536,0.014855111638704935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,16384,0.05335022343529595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,768,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,65536,0.12219644917382134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,51200,0.13025511635674372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,12288,0.043346666627460055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,16384,0.04632444514168633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,16384,0.035754667388068304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,10240,0.03526399864090814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,12288,0.03532888823085361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,12288,0.0281315545241038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,16384,0.05061422122849358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,8192,0.029879109727011785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,10240,0.03072355522049798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,12288,0.04064977831310696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,65536,0.16029243999057347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,7168,0.02566844390498267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,8192,0.02687377731005351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,8192,0.020958221620983545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,10240,0.035912000470691256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,6144,0.022291556000709534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,7168,0.01907199952337477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,8192,0.031694223483403526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,6144,0.020280000236299302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,6144,0.01719466679626041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,7168,0.028335111008750066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,5120,0.019309333629078336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,51200,0.09686044189665054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,5120,0.015795555379655626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,6144,0.026000888811217412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,4096,0.01641688909795549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,5120,0.017621333400408428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,5120,0.02343911098109351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,51200,0.12778845098283556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,10240,0.024544888072543677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,3584,0.014202665951516895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,3584,0.01331822243001726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,3072,0.012392889294359418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,3584,0.014212444424629211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,7168,0.02295111119747162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,3584,0.019476445184813607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,2560,0.010433778166770935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,3072,0.013064888616402945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,3072,0.012259555359681448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,3072,0.017887110511461895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,2560,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,2048,0.008935111264387766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,2560,0.01735911104414198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,2048,0.010836444795131683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,1536,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,4096,0.014004444082578024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,2048,0.009453333086437648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,2048,0.015822221835454304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,1024,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,1536,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,4096,0.02087200019094679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,1536,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,768,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,1024,0.007673777639865875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,1536,0.014911999305089315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,512,0.0037137779096762338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,768,0.007330666813585494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,768,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,1024,0.012322666744391123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,256,0.0030373332815037835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,512,0.006817777951558431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,768,0.011415111521879831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,2560,0.011741333537631564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,128,0.0028106667515304354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,256,0.006555555595291986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,256,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,512,0.011023999916182624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,64,0.002584888910253843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,128,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,128,0.00657066661450598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,4096,0.015478221906556023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,32,0.002854222224818336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,256,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,64,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,128,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,65536,0.18142310778299967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,32,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,1024,0.007671111159854465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,65536,0.1278266641828749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,51200,0.14278133710225424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,16384,0.04741866721047295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,65536,0.07272089189953275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,51200,0.0973857773674859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,51200,0.058232002788119845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,12288,0.03734044565094842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,512,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,16384,0.023358222511079576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,16384,0.035192890299691096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,10240,0.0314977765083313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,12288,0.01866044435236189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,12288,0.027789334456125896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,65536,0.14385688304901123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,16384,0.04622222317589653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,8192,0.027136888768937852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,51200,0.11547288629743789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,10240,0.02381688853104909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,7168,0.02291822267903222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,12288,0.03801866703563266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,8192,0.01482488877243466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,6144,0.01997777819633484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,8192,0.020245333512624104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,10240,0.033655110332700945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,7168,0.01351199961370892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,5120,0.017139555679427255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,7168,0.017839110559887357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,6144,0.012711111042234631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,8192,0.029454221328099568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,6144,0.01644088824590047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,4096,0.014669333895047506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,7168,0.02733777794573042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,5120,0.011490667031870948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,5120,0.014365333649847241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,3584,0.011694221860832639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,6144,0.024313777685165405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,4096,0.010658666491508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,4096,0.01314311143424776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,3072,0.010653333531485664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,3584,0.0099973330895106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,5120,0.021977777282396953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,3584,0.011904000408119626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,2560,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,4096,0.019652444455358718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,3072,0.008864888714419471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,2048,0.00795199970404307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,3072,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,3584,0.01891377733813392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,2560,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,2560,0.010093332992659675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,10240,0.016505777835845947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,1536,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,2048,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,2048,0.009293333523803288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,2560,0.016726222303178575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,1536,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,1536,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,1024,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,768,0.004238222208287981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,1536,0.013789333403110504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,1024,0.011413333316644033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,768,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,512,0.003549333247873518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,768,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,768,0.01071200023094813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,512,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,512,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,3072,0.017557332913080852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,512,0.00998933364947637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,256,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,2048,0.01516177753607432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,256,0.005889777921968036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,128,0.0028239999794297745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,256,0.009675555759006077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,1024,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,128,0.00609777785009808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,32,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,128,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,128,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,64,0.0058408888677755995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,65536,0.15983200073242188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,32,0.006224000205596288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,51200,0.1272053321202596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,256,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,65536,0.12372533480326335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,65536,0.06896533568700154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,16384,0.044590221510993115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,1024,0.005535111245181825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,51200,0.09485777881410386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,51200,0.056143999099731445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,12288,0.03525866733656989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,16384,0.03512533174620734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,64,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,16384,0.021976888179779053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,10240,0.028677334388097126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,16384,0.045019557078679405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,12288,0.01797422269980113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,12288,0.027104000250498455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,8192,0.024941333466105994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,10240,0.015948444604873657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,12288,0.037575112448798284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,7168,0.02146311104297638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,8192,0.020077332854270935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,10240,0.03333866596221924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,8192,0.02921244502067566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,7168,0.01759822169939677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,6144,0.018516444497638278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,7168,0.026143110460705225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,6144,0.016138666205936007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,5120,0.01593955523437924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,6144,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,6144,0.02415200074513753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,65536,0.14201422532399496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,4096,0.013648889131016202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,5120,0.014147554834683737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,51200,0.11438488960266113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,10240,0.022766222556432087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,4096,0.012867555850081973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,4096,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,3584,0.011045333411958484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,5120,0.021456888980335657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,8192,0.014114666316244336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,3072,0.009999111294746399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,3584,0.00907199995385276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,4096,0.019715555840068393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,3584,0.011794666449228922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,2560,0.008745777938101027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,7168,0.013106666505336761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,3072,0.010946667028797997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,2048,0.007615110940403408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,3584,0.018593778212865193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,2560,0.00777422222826216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,3072,0.017290666699409485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,2560,0.010019555687904358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,1536,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,2048,0.007328888608349695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,2048,0.00847111145655314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,2560,0.016136889656384785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,1024,0.004690666579537922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,1536,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,1536,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,2048,0.014834667245546976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,768,0.003875555677546395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,5120,0.011387555963463254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,1024,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,1024,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,512,0.0034888887570963967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,768,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,768,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,1024,0.011339555184046427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,256,0.0031582222630580268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,512,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,512,0.006119110931952794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,768,0.01110666659143236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,128,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,256,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,512,0.010513777534166971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,128,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,256,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,128,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,32,0.0028373334142896864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,128,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,64,0.005631111148330901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,32,0.006080889039569431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,65536,0.14432266023423937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,1536,0.012506666282812754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,51200,0.11397243870629205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,65536,0.11876711580488418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,65536,0.06709422005547418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,3072,0.008231999973456064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,51200,0.09001866976420085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,16384,0.0348417792055342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,51200,0.05468711256980896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,65536,0.14230755964914957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,64,0.002516444358560774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,12288,0.02790400054719713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,16384,0.021311110920376245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,16384,0.03326311045222812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,51200,0.11372978157467312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,10240,0.023319111929999456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,12288,0.025821333130200703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,12288,0.01755466726091173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,8192,0.02033955521053738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,16384,0.04531288809246487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,10240,0.022258667482270136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,10240,0.015493333339691162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,12288,0.03730666637420654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,7168,0.017140444782045152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,8192,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,8192,0.01902577777703603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,10240,0.03292266527811686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,6144,0.014864888456132678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,7168,0.012748444245921241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,7168,0.017120889491505094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,8192,0.027857777145173814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,5120,0.013072888884279462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,7168,0.026168000366952684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,6144,0.015588444140222339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,4096,0.011023999916182624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,5120,0.010692444112565784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,5120,0.013806222213639153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,3584,0.010025777750545079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,4096,0.009375111096435124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,4096,0.012477333347002665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,5120,0.021512000097168818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,3072,0.008599999878141616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,3584,0.008630221916569604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,3584,0.018403554956118267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,3072,0.01071466671095954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,2560,0.007816889219813878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,3072,0.008369777765538957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,3072,0.017088888419999015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,256,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,2560,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,2048,0.0069013333155049225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,2560,0.0076044441925154785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,6144,0.011744889120260874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,1536,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,2048,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,2560,0.016157334049542744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,6144,0.023801777097913954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,1024,0.004559999952713649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,2048,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,1536,0.00665866666369968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,1536,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,768,0.003540444291300244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,1024,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,3584,0.01144977741771274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,1536,0.0123831108212471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,512,0.0031715554909573663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,768,0.005971555494599872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,768,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,1024,0.011160888605647616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,256,0.0028071111689011254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,512,0.005977777971161737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,512,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,768,0.01053600013256073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,128,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,256,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,256,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,512,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,64,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,128,0.0052915554907586836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,256,0.00963199966483646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,32,0.00253866674999396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,65536,0.11760621600680882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,64,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,128,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,2048,0.014491554763582019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,32,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,4096,0.01941511034965515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,1024,0.006568000134494569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,51200,0.09630044301350911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,65536,0.11083110835817124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,16384,0.03763466742303636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,51200,0.08691022131178115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,51200,0.05206933286454943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,16384,0.020231111182106864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,16384,0.03194844391610887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,65536,0.14142933156755236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,12288,0.029584887954923842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,128,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,10240,0.021016889148288306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,16384,0.044785777727762856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,12288,0.01657511128319634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,12288,0.024487111303541396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,8192,0.018108444081412423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,10240,0.015015999476114908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,10240,0.02166844407717387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,12288,0.036511109934912786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,7168,0.015536889433860779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,8192,0.018911111685964797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,10240,0.0316880014207628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,65536,0.06418577829996745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,6144,0.013694222602579327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,8192,0.027129777603679236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,7168,0.012401777837011548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,7168,0.016699555847379897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,5120,0.011811555259757571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,6144,0.01148977792925305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,7168,0.025242666403452556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,4096,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,51200,0.1134035587310791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,5120,0.010068444742096795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,5120,0.013384000294738345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,3584,0.009285333255926767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,6144,0.023381332556406658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,4096,0.01219911128282547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,4096,0.008709333009190029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,5120,0.021031111478805542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,3072,0.008107555409272512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,3584,0.008683555656009251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,4096,0.019260444574885897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,2560,0.007264888948864407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,3072,0.010339555641015371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,3584,0.018043556147151522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,8192,0.013407111167907715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,2048,0.006540444576077991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,3072,0.01685066686736213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,2560,0.007703999678293864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,2560,0.009015999734401703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,1536,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,2048,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,2048,0.007720000214046902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,2560,0.014888000157144336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,6144,0.015289778510729471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,1024,0.004152889053026835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,1536,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,1536,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,768,0.0035413333939181436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,1024,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,1024,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,1536,0.01275644451379776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,512,0.0031973332580592898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,768,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,3584,0.011386666860845355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,1024,0.01107377807299296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,256,0.002809777855873108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,512,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,768,0.011348444554540845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,512,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,3072,0.008076444268226624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,128,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,256,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,256,0.00553955551650789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,512,0.009663999908500248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,64,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,128,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,128,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,256,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,128,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,65536,0.09138755665885077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,64,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,32,0.005640000104904175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,2048,0.01480888823668162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,51200,0.07686755392286512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,65536,0.06960622469584148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,65536,0.06174577607048882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,16384,0.029951112137900457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,51200,0.0551128884156545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,51200,0.0499271121289995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,768,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,12288,0.023927110764715407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,65536,0.14104533195495605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,16384,0.021141333712471858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,10240,0.01812622282240126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,51200,0.11313954989115398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,12288,0.016925333274735343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,12288,0.01639111174477471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,16384,0.044383111927244395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,8192,0.015146666102939181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,10240,0.015046222342385186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,32,0.002455111179086897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,12288,0.03543377916018168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,7168,0.013415111435784234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,8192,0.013050666285885705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,10240,0.031000001562966242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,7168,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,7168,0.01183111137813992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,6144,0.011935999823941125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,8192,0.027292443646325007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,6144,0.010978666444619497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,7168,0.024838222397698298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,5120,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,6144,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,16384,0.019655111763212416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,4096,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,5120,0.009655999640623728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,5120,0.010315555665228102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,5120,0.020925333102544148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,4096,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,4096,0.008669333325492011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,10240,0.014536889062987434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,4096,0.0185217774576611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,3584,0.008019555773999956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,3072,0.007255111303594377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,8192,0.013109332985348172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,3584,0.008053333395057255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,2560,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,3072,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,3584,0.016919111212094624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,3072,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,2048,0.005700444595681296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,2560,0.006989333364698622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,3072,0.015638222297032673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,2048,0.006548444430033366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,2048,0.007035555938879649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,1536,0.004416000097990036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,2560,0.014659555421935188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,2048,0.01313244468635983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,1536,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,1024,0.003539555602603488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,1536,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,3584,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,768,0.0031911111954185697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,1536,0.012133333418104382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,1024,0.0063288890653186375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,512,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,1024,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,768,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,768,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,256,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,512,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,768,0.01109688894616233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,512,0.01035733355416192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,256,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,256,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,128,0.002463999928699599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,128,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,6144,0.022680888573328655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,64,0.0022373333987262514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,256,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,32,0.002463111033042272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,128,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,65536,0.0810862249798245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,128,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,32,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,65536,0.06798222329881456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,51200,0.06810577710469563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,65536,0.06090488698747423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,1024,0.01075644459989336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,51200,0.052631109952926636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,16384,0.027740443746248882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,51200,0.0488026671939426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,512,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,65536,0.14113067256079778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,12288,0.0226248883538776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,16384,0.019561777512232464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,2560,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,51200,0.1127742264005873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,10240,0.019724445210562814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,16384,0.020484444167878892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,12288,0.01607111096382141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,8192,0.014764444695578681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,16384,0.043713778257369995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,10240,0.014404444230927361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,12288,0.03530844383769565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,10240,0.014487110906177096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,7168,0.01275644451379776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,8192,0.012522666818565793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,8192,0.013075555364290873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,64,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,6144,0.011347555451922946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,10240,0.03107822272512648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,7168,0.01181688904762268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,5120,0.010029333333174387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,8192,0.02681066592534383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,6144,0.010145777629481422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,7168,0.024597333537207708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,5120,0.00997866690158844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,5120,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,4096,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,6144,0.02289955483542548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,4096,0.008634666601816813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,5120,0.02068711154990726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,3584,0.007768889268239339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,4096,0.008971555365456475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,12288,0.01633777717749278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,3072,0.007225777539942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,4096,0.01831999917825063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,3584,0.008030222521887885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,3584,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,2560,0.006439111298984951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,3072,0.007704888780911763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,3072,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,7168,0.011167999770906238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,2048,0.005471999860472149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,2560,0.007535999847782983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,2560,0.006903111106819577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,6144,0.011235555840863122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,1536,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,2560,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,2048,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,2048,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,1024,0.0035377778112888336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,1536,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,2048,0.013439111411571503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,1536,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,768,0.003157333367400699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,1024,0.00564533347884814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,1024,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,1536,0.012098666694429187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,512,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,768,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,1024,0.011011555790901184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,512,0.005295111073387994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,3584,0.017847999930381775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,256,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,512,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,3072,0.01516888870133294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,512,0.010015111002657149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,128,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,256,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,256,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,128,0.00508266646001074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,128,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,32,0.0022764443937275144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,64,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,65536,0.0760177771250407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,768,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,32,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,51200,0.06219644678963555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,65536,0.065975997183058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,768,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,65536,0.05952711237801445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,16384,0.023686221904224817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,256,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,65536,0.14073155985938177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,51200,0.04785244332419502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,12288,0.020579554968410067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,16384,0.019488000207477145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,16384,0.01926577753490872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,128,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,64,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,10240,0.016322667400042217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,51200,0.11253778139750163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,12288,0.015823110938072205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,8192,0.0148044443792767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,16384,0.04336355461014641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,10240,0.01386666629049513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,10240,0.014470222923490735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,12288,0.035142223040262856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,7168,0.012587555580668978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,8192,0.011789333489206104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,8192,0.01253600004646513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,6144,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,7168,0.011008888483047485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,7168,0.011579555769761404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,8192,0.026727999250094097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,5120,0.009604444106419882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,7168,0.02444800072246128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,6144,0.010772444307804108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,51200,0.051073776351081006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,4096,0.008560000194443597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,5120,0.009116444322797988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,6144,0.02236000034544203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,5120,0.009705777797434065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,3584,0.007530666887760162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,5120,0.020269332660569083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,4096,0.008092444803979661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,12288,0.01588088936275906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,3072,0.007059555914666917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,3584,0.008006222546100616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,4096,0.017874666386180453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,3584,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,2560,0.006208000083764394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,3072,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,3072,0.007837333612971837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,10240,0.030559112628300984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,3584,0.016539555456903245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,2048,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,2560,0.007330666813585494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,2560,0.006913777854707506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,3072,0.0151573336786694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,1536,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,6144,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,2048,0.0064151109092765385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,2048,0.007227555745177799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,2560,0.014372444815105863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,1024,0.003531555748648114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,1536,0.006745777610275481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,1536,0.00612088872326745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,768,0.0033982222278912864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,2048,0.013761777844693927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,1024,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,1024,0.005643555687533484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,4096,0.008375999828179678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,512,0.0031351111829280853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,1536,0.012040889097584618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,768,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,768,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,1024,0.011050666371981302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,256,0.002510222295920054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,512,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,128,0.0024764444679021835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,768,0.010838222172525195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,512,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,512,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,256,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,64,0.0022168887986077205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,128,0.005677333308590784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,128,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,256,0.009339555270142024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,65536,0.06180266539255778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,64,0.004983110974232356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,128,0.007799111306667328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,32,0.005294222384691238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,51200,0.051728000243504844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,65536,0.06311555703481038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,65536,0.05834844377305773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,16384,0.021242666575643752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,51200,0.0490497781170739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,51200,0.04641955428653293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,12288,0.017638223038779367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,65536,0.14023999373118082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,16384,0.01921688848071628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,16384,0.01889866590499878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,10240,0.015269334117571512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,51200,0.11275377538469102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,12288,0.015389333168665567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,256,0.005710222240951326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,12288,0.015588444140222339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,16384,0.042527112695905894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,8192,0.014119111829333834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,10240,0.013457777599493662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,10240,0.013739555246300168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,32,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,12288,0.03475111060672336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,7168,0.01239999963177575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,8192,0.011424889167149862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,6144,0.011357333097192975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,8192,0.012183999849690331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,10240,0.030576888057920668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,7168,0.010616888602574667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,5120,0.009641778137948778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,7168,0.01126844435930252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,6144,0.010405333505736457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,4096,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,7168,0.024858666790856257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,5120,0.009000000026490953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,5120,0.009199110998047723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,6144,0.02240444388654497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,3584,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,4096,0.00831288927131229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,4096,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,3072,0.006704000135262807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,3584,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,4096,0.01768977774514092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,3584,0.016181333197487723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,3072,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,2560,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,3072,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,2560,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,3072,0.0151555554734336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,2560,0.006793777975771163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,2048,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,6144,0.009881777895821465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,2560,0.014088888963063559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,2048,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,1536,0.00425511101881663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,2048,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,1024,0.0035111109415690103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,1536,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,2048,0.01312355531586541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,1536,0.006196444647179709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,768,0.0031840000301599503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,5120,0.020642666353119742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,1024,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,1024,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,3584,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,512,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,768,0.006227555374304454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,768,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,512,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,512,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,768,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,512,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,256,0.004955555415815777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,128,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,8192,0.026584888497988384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,256,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,256,0.009319110876984065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,64,0.0021848889688650766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,128,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,32,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,128,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,65536,0.05381066600481669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,1536,0.012415111064910889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,128,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,64,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,32,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,1024,0.01106844428512785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,65536,0.056271996763017446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,65536,0.06016355752944946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,16384,0.017384888397322763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,256,0.002495999965402815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,51200,0.046887109676996865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,51200,0.045392000012927584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,65536,0.14080088668399388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,12288,0.01387466655837165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,16384,0.018570666511853535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,51200,0.1120835542678833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,10240,0.012712889247470431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,12288,0.014787554740905762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,12288,0.015189333094490899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,16384,0.04252622193760342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,8192,0.012773333324326409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,10240,0.012792000340090858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,10240,0.013470222552617392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,12288,0.03449511196878221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,7168,0.011558222273985544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,8192,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,8192,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,10240,0.030293332205878362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,51200,0.042509333954917065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,6144,0.010272000398900773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,7168,0.01056977775361803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,7168,0.011056889262464313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,5120,0.009110222260157267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,6144,0.009707555174827576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,6144,0.009679111341635386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,4096,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,7168,0.02421066661675771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,5120,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,6144,0.021711111068725586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,5120,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,3584,0.007069333559936947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,16384,0.018555555078718398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,4096,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,3072,0.006510222123728857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,3584,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,5120,0.019498666127522785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,3584,0.0075786668393347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,4096,0.016941333810488384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,2560,0.0057608890864584185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,3072,0.007673777639865875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,3072,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,3584,0.01594577729701996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,2048,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,2560,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,2560,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,3072,0.015104888214005364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,1536,0.003924444317817688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,2048,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,2048,0.006311111152172089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,1024,0.0032071111102898917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,2560,0.013906665974193148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,8192,0.026250667042202417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,1536,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,1536,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,2048,0.012783110969596438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,1024,0.006330666856633292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,1024,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,1536,0.011758222348160215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,512,0.0025902222841978073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,768,0.006014222072230444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,1024,0.011015111373530494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,4096,0.007876444194051955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,256,0.002248888835310936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,768,0.010406222608354356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,512,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,128,0.0021297778520319196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,512,0.009668444593747457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,256,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,64,0.0019164443429973389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,128,0.0058328890138202244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,128,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,256,0.009539555344316695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,32,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,65536,0.04009155432383219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,64,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,128,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,768,0.0028817777832349143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,51200,0.03314400050375197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,65536,0.05730311075846354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,65536,0.0557546681827969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,768,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,51200,0.04475111100408766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,16384,0.014077333940400017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,51200,0.04549955659442478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,512,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,12288,0.012442666623327466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,16384,0.017494221528371174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,16384,0.018218666315078735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,51200,0.11211733023325603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,10240,0.011015111373530494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,12288,0.01480977733929952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,16384,0.04216622312863668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,12288,0.013461333182122974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,8192,0.00999022192425198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,10240,0.012151111331250934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,12288,0.03461688756942749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,32,0.005199999858935674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,8192,0.01107377807299296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,8192,0.011015111373530494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,10240,0.030461334519916113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,7168,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,8192,0.025399110383457605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,6144,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,65536,0.14020000563727483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,7168,0.02402399977048238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,6144,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,5120,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,6144,0.009295110901196798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,5120,0.008974221845467886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,6144,0.02165422174665663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,4096,0.007705777883529663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,5120,0.008657777474986183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,5120,0.019559999306996662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,4096,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,10240,0.012897777888509961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,3584,0.006568000134494569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,4096,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,3072,0.005975111077229182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,3584,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,4096,0.016864889197879367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,3584,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,3072,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,3072,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,3584,0.015804444750150044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,7168,0.010044444766309526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,2048,0.00453599997692638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,2560,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,3072,0.014766222900814481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,1536,0.0041777777175108595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,2048,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,2048,0.006424888968467712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,2560,0.013853333062595792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,1024,0.00328177772462368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,1536,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,2048,0.012895111408498553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,1536,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,768,0.0032026666320032547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,1024,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,1024,0.005728888842794631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,1536,0.0120666664507654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,512,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,768,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,768,0.005986666513813867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,1024,0.010689777632554373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,7168,0.009341333475377824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,256,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,2560,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,512,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,512,0.0052933332820733385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,128,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,768,0.010460444622569613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,256,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,64,0.0021199999998013177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,256,0.005199111170238919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,512,0.009672000176376766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,32,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,2560,0.006859555426571105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,128,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,256,0.008952000074916417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,64,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,128,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,32,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,51200,0.024096000525686476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,65536,0.055009779002931386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,16384,0.011569778124491373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,51200,0.04417599903212654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,51200,0.04378488990995619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,65536,0.1401537789238824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,16384,0.017539555827776592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,12288,0.01258666647805108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,51200,0.11133955584632026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,16384,0.016925333274735343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,16384,0.0420648886097802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,12288,0.013415999710559845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,12288,0.013794666363133324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,10240,0.011153777440388998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,8192,0.010107555323176915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,10240,0.012067555553383298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,12288,0.03398222062322829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,65536,0.03103644318050808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,7168,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,10240,0.012270222107569376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,8192,0.010835555692513784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,10240,0.02947910957866245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,65536,0.05506044295099047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,6144,0.008656000097592672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,7168,0.010064000056849586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,7168,0.01034044474363327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,8192,0.025044444534513686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,5120,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,6144,0.009737778041097853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,7168,0.02332088847955068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,6144,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,4096,0.006996444529957241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,5120,0.008861333131790161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,6144,0.020941333638297185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,3584,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,4096,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,4096,0.008319999608728621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,5120,0.01924444403913286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,3072,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,4096,0.01720622181892395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,3584,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,3072,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,3072,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,3584,0.015450666348139444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,8192,0.010743111371994019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,3072,0.014838222000333997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,2048,0.004551110996140374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,2560,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,2560,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,2560,0.013752889302041797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,1536,0.003887999802827835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,2048,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,2048,0.006317333214812809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,5120,0.008535111116038429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,1024,0.0032008888406885993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,1536,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,1536,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,768,0.0028728888266616394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,3584,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,1024,0.006320000108745363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,1536,0.011706666813956367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,512,0.002483555633160803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,768,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,2560,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,1024,0.011032889286677042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,256,0.002198222196764416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,512,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,512,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,768,0.010393777655230628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,128,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,256,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,256,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,512,0.009375111096435124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,64,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,128,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,128,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,32,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,256,0.008815999660227034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,65536,0.02537511123551263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,64,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,2048,0.012737777498033313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,32,0.004997333304749595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,51200,0.020406223005718656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,65536,0.0536426670021481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,65536,0.05474488933881124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,1024,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,16384,0.01238666640387641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,768,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,51200,0.04297688934538099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,51200,0.04386666748258802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,12288,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,16384,0.01682133310370975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,16384,0.016232000456915963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,10240,0.009708444277445475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,51200,0.11138755745357937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,12288,0.013425778183672162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,12288,0.01348444405529234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,8192,0.00888355573018392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,10240,0.012081777883900536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,12288,0.03325244453218248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,10240,0.012114666402339935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,7168,0.00832088871134652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,8192,0.010999999940395355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,8192,0.010614222122563256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,10240,0.02947288751602173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,6144,0.007847111258241866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,7168,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,7168,0.01032977799574534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,8192,0.02498133314980401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,6144,0.009376000199053023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,6144,0.009640889035330879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,5120,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,65536,0.1399946610132853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,6144,0.020953777763578627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,4096,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,5120,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,5120,0.008616888688670265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,16384,0.04177244504292806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,3584,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,4096,0.008271999657154083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,5120,0.019181332654423185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,3584,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,4096,0.017157332764731515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,3072,0.005543111099137201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,128,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,3584,0.016121778223249648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,2560,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,3072,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,3072,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,7168,0.022655111220147874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,3072,0.014768888552983602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,2048,0.004519999855094486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,2560,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,2048,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,2560,0.014095111025704278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,1536,0.0038240001433425476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,2048,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,1024,0.0031822222388452957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,1536,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,2048,0.012800888882742988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,4096,0.007715555528799693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,1536,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,768,0.0028604444944196274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,1024,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,1024,0.006329777754015393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,3584,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,512,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,1536,0.012062222593360476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,768,0.006092444476154115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,768,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,256,0.002193777718477779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,512,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,1024,0.011048888994587792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,128,0.002033777741922273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,768,0.010085333552625444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,512,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,256,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,64,0.0018035556293196147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,256,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,512,0.00980355590581894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,768,32,0.0018177777528762817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,2560,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,128,0.005648889061477449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,128,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,256,0.008948444492287105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,51200,0.01564088960488637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,64,0.005190222213665645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,32,0.004964444372389051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,128,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,65536,0.052516443861855396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,16384,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,51200,0.04040266738997565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,51200,0.042358222934934825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,12288,0.008745777938101027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,65536,0.13935911655426025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,16384,0.016389333539538913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,16384,0.016187555260128446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,10240,0.00777599960565567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,51200,0.11149688561757405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,16384,0.04091822107632955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,12288,0.013440888788965015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,12288,0.013452444639470843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,8192,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,10240,0.012091555529170565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,10240,0.012075555821259817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,7168,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,12288,0.033089776833852134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,8192,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,10240,0.029135998752382066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,65536,0.01926399932967292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,6144,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,8192,0.024719999896155462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,7168,0.01019022199842665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,7168,0.010018666585286459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,5120,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,65536,0.053731554084353976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,6144,0.00904088881280687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,4096,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,7168,0.022700443863868713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,5120,0.008920888933870528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,5120,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,3584,0.005983110931184556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,6144,0.020971554848882887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,5120,0.018580444984965853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,4096,0.00835111074977451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,4096,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,3072,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,3584,0.007400888535711501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,4096,0.016864000095261466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,3072,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,3584,0.016196444630622864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,2560,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,3072,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,8192,0.010341333018408881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,2048,0.004242666479614046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,2560,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,3072,0.0151591118839052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,2560,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,1536,0.0038559999730851916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,2048,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,2560,0.013888888888888888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,6144,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,2048,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,1024,0.003216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,1536,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,1536,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,768,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,2048,0.012423999607563019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,1024,0.006303999986913469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,1024,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,512,0.0024862223201327855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,1536,0.012422222230169507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,768,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,768,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,3584,0.008005333443482717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,256,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,512,0.0058755555914507965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,512,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,768,0.010448888772063784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,256,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,512,0.010037333601050908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,256,0.009687999884287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,128,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,64,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,32,0.0018177777528762817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,128,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,65536,0.015218666858143277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,64,0.005103111267089844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,32,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,65536,0.047320889102088086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,65536,0.05173155665397644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,51200,0.013236444857385425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,1024,0.010684444672531553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,65536,0.1385857794019911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,51200,0.03803911142879062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,512,128,0.0021235555824306277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,51200,0.039841777748531766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,256,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,16384,0.015967110792795818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,51200,0.1104302207628886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,16384,0.016366221838527255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,10240,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,12288,0.013471110827393003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,12288,0.013429332938459186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,16384,0.04073066512743632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,8192,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,10240,0.012064889073371887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,12288,0.032856888241238065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,10240,0.011855111353927188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,7168,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,8192,0.010390222072601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,8192,0.010687111152542962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,10240,0.02846222122510274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,6144,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,7168,0.010098666780524785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,7168,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,8192,0.024910221497217815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,5120,0.006535111202134027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,16384,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,6144,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,6144,0.009655111365848118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,7168,0.022351110974947613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,12288,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,5120,0.008843555218643611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,5120,0.008357333640257517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,6144,0.020595555504163105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,3584,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,128,0.0057760001056724125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,4096,0.008263111114501953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,4096,0.007709333466158972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,3072,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,5120,0.01905955539809333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,3584,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,2560,0.004958222309748332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,4096,0.01684177749686771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,3584,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,3072,0.007612444460391998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,3072,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,2048,0.004687110996908612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,2560,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,3584,0.016192888220151264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,1536,0.0035208890007601846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,2560,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,3072,0.014482667048772177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,2048,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,1024,0.003149333306484752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,2048,0.006218666831652324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,2560,0.013787555197874704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,1536,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,1536,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,2048,0.01311377767059538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,1024,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,1024,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,4096,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,1536,0.012418666647540199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,512,0.0024622221373849446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,768,0.005970666805903117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,1024,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,768,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,256,0.0021297778520319196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,512,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,768,0.010062221851613786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,256,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,128,0.001863111017478837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,512,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,256,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,64,0.001826666709449556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,128,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,32,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,128,0.007258666886223688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,64,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,256,768,0.0028124445428450904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,32,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,65536,0.049294223388036094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,51200,0.010478222535716163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,512,0.0057751110030545135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,65536,0.13784800635443792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,51200,0.03962755534383986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,256,0.0057502223385704895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,16384,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,51200,0.03845244314935472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,128,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,12288,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,16384,0.015976000163290236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,16384,0.016166221764352586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,51200,0.10975199937820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,12288,0.013431111143694984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,12288,0.013449778159459433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,10240,0.011762667033407422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,12288,0.032551109790802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,10240,0.012072888513406118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,8192,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,65536,0.04656177759170532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,7168,0.005378666851255629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,8192,0.010673777924643623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,10240,0.029115554359224107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,8192,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,6144,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,7168,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,7168,0.00980888886584176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,8192,0.024668445189793903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,5120,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,6144,0.009009777671760982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,7168,0.02250755495495266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,5120,0.008743111458089616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,5120,0.008276444342401292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,10240,0.0058017778727743365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,4096,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,6144,0.02128622266981337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,3584,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,65536,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,4096,0.008298666940795051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,5120,0.018578666779730055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,3072,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,4096,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,3584,0.007956444389290279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,3584,0.00738666703303655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,2560,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,4096,0.017208000024159748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,3072,0.007551111280918121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,3584,0.015513777732849121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,3072,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,2048,0.004587555511130227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,2560,0.007270221908887227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,2560,0.00665155549844106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,3072,0.01457244488928053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,2048,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,2048,0.006060444646411472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,2560,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,1024,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,6144,0.009463110731707679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,2048,0.01274311128589842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,1536,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,768,0.002622222320901023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,1024,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,1024,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,16384,0.04073866539531284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,1536,0.011708444191349877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,512,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,1024,0.011371555427710215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,768,0.006065777606434292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,768,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,256,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,512,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,512,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,128,0.0018035556293196147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,256,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,256,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,256,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,128,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,64,0.0018195555441909367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,128,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,32,0.001826666709449556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,128,0.007176000210973952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,1536,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,65536,0.00851999968290329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,64,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,51200,0.007709333466158972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,16384,0.005712000032265981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,32,0.004970666848950916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,12288,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,65536,0.04862844281726413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,51200,0.03924444317817688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,16384,0.015933333171738517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,8192,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,12288,0.013453333742088742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,7168,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,768,0.010061333576838175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,10240,0.011703111231327057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,6144,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,512,0.010007999837398529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,5120,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,4096,0.004611555486917496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,8192,0.010317333042621613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,3584,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,7168,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,128,1536,0.0035075553589397003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,3072,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,6144,0.009126221968068017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,5120,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,2048,0.004570666700601578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,4096,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,1536,0.003479111111826367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,3584,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,3072,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,1024,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,768,0.0025235555238193935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,2560,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,512,0.002472888885272874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,2048,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,256,0.0021226666867733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,1536,0.0057884446448749965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,1024,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,64,0.0018053334206342697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,10240,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,768,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,32,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,256,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,65536,0.011095999843544431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,51200,0.009720000127951304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,128,0.004599110947714912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,16384,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,64,0.005069333232111401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,32,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,65536,0.04902311166127523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,51200,0.03907555672857497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,16384,0.01591911084122128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,8192,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,2560,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,12288,0.013096888860066732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,7168,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,6144,0.005019555489222209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,10240,0.011756444142924415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,5120,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,8192,0.010070222119490305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,4096,0.004518222063779831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,7168,0.009655999640623728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,3584,0.004593777573770947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,64,128,0.001855111163523462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,3072,0.004641777939266629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,4096,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,512,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,2560,0.004862222406599256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,2048,0.00452533322903845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,3584,0.007132444116804335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,1536,0.0034959999223550162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,3072,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,1024,0.0028533333291610083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,2560,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,12288,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,768,0.0025439999169773525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,2048,0.005956444475385878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,512,0.0024862223201327855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,256,0.0021297778520319196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,10240,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,1536,0.005752000129885144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,128,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,64,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,768,0.005335999859703912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,512,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,128,32,32,0.001793777777089013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,256,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,128,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,64,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,32,0.004603555632962121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,6144,0.009333333207501305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,5120,0.00813244448767768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,16384,0.3723111152648926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,16384,0.2044853369394938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,16384,0.13363556067148843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,51200,0.6254968643188477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,1024,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,51200,1.141858630710178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,10240,0.23782221476236978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,12288,0.10015644629796346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,8192,0.19476977984110513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,10240,0.12720799446105957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,10240,0.0860862202114529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,51200,0.40265599886576336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,7168,0.16883022255367705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,8192,0.11119200123680963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,6144,0.14490044116973877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,12288,0.28090045187208385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,7168,0.06314844555324979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,6144,0.08099466562271118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,8192,0.07141511307822333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,5120,0.12217599815792507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,12288,0.150255110528734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,6144,0.05362400081422594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,5120,0.05118044548564487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,3584,0.08749155865775214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,5120,0.0675191084543864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,4096,0.05626222160127428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,3072,0.07684622208277385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,3584,0.05012977785534329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,3584,0.03546488947338528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,2560,0.06396977769003974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,3072,0.04316355453597175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,3072,0.0311173333062066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,2048,0.051595555411444775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,1536,0.04051999913321601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,2560,0.02875111169285244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,2048,0.023595554961098567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,2560,0.03768710957633124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,1024,0.028971556160185073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,2048,0.032467554012934365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,768,0.021698666943444148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,1536,0.018974221414989896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,1024,0.016138666205936007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,512,0.015068444940778943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,1024,0.020837333467271592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,1536,0.025586666332350835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,7168,0.09634666972690159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,256,0.009658666948477427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,4096,0.09788533051808675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,512,0.012716444830099741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,128,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,256,0.011712888876597086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,512,0.013642667068375481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,64,0.004987555659479565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,256,0.012423999607563019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,32,0.0051893335249688895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,128,0.011401777466138205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,128,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,4096,0.03921688927544488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,64,0.011358222199810876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,32,0.015196444259749519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,65536,0.6322435802883571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,768,0.014428445034556918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,51200,0.49246576097276473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,768,0.016179554992251925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,65536,1.1179822285970051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,16384,0.28412532806396484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,12288,0.2153280046251085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,16384,0.1616213321685791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,16384,0.10734755463070339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,51200,0.3191893365648058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,10240,0.16693422529432508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,65536,0.4032622178395589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,8192,0.141047994295756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,12288,0.1196862194273207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,10240,0.06796089145872328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,10240,0.10149333212110732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,7168,0.11919732888539632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,8192,0.08605421913994683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,6144,0.10301422410541111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,8192,0.05861688984764946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,7168,0.07275911172231038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,7168,0.04996444450484382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,5120,0.08744711346096462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,4096,0.07318133115768433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,6144,0.06406488683488634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,6144,0.043600890371534556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,3584,0.06297866503397624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,5120,0.05508888761202494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,3072,0.05470222234725952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,3584,0.02938133478164673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,4096,0.03571022219128079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,3584,0.04023733403947618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,4096,0.044638223118252225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,3072,0.025763556361198425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,3072,0.03517244590653314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,12288,0.08153155777189466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,2560,0.030968000491460163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,2048,0.03806844353675842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,1536,0.029785778787400987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,2048,0.019208888212839764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,2048,0.02700888944996728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,1536,0.01612711118327247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,1024,0.02239377796649933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,1536,0.021280889709790547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,768,0.016696888539526198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,51200,0.8769804636637369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,1024,0.01447111037042406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,5120,0.03712711069318984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,768,0.013756444056828817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,512,0.011944889194435544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,2560,0.046720888879564076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,256,0.007791111038790808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,256,0.010050666828950247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,2560,0.023919110496838886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,128,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,256,0.010768888725174798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,64,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,128,0.010336000058386061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,32,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,128,0.00999111102686988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,64,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,32,0.012847111456924014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,1024,0.016772444049517315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,65536,0.32452000512017143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,65536,0.22958932982550728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,51200,0.2515795495775011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,512,0.011413333316644033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,16384,0.0852560003598531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,65536,0.14062310589684382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,51200,0.17527822653452554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,12288,0.06552088922924466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,16384,0.06077422036064995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,51200,0.11271111170450847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,16384,0.040968888335757785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,10240,0.054442667298846774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,8192,0.045239110787709556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,12288,0.04632888899909126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,12288,0.03305600086847941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,10240,0.027930667002995808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,7168,0.039932443035973444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,10240,0.03987733191914029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,8192,0.02372444503837162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,6144,0.03502310978041755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,512,0.011375111010339526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,5120,0.029822223716311987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,7168,0.030135999123255413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,6144,0.019455111689037748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,4096,0.025436444414986506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,6144,0.02660977840423584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,5120,0.017584888471497428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,3584,0.021751999855041504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,5120,0.023008000519540574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,768,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,4096,0.01645066671901279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,3072,0.019301333361201815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,2560,0.016216889023780823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,3584,0.014114666316244336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,4096,0.019621334142155118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,3584,0.017376888129446242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,2048,0.0136835558546914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,3072,0.013188444077968597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,1536,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,3072,0.015825778245925903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,2048,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,2560,0.013912889692518445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,1024,0.008628444539176093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,1536,0.009345778160625035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,2048,0.012479110724396177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,768,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,1536,0.010784000158309937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,512,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,1024,0.008569777839713627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,1024,0.009525333013799455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,768,0.007744889292452071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,256,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,768,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,128,0.003154666680428717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,7168,0.02128533356719547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,512,0.007043555378913879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,256,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,64,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,32,0.0029226665695508323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,128,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,256,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,128,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,64,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,32,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,65536,0.24296177758110893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,2560,0.012103999654452005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,65536,0.17697778013017443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,16384,0.06532710790634155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,51200,0.13509599367777506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,12288,0.04901955525080363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,16384,0.03748977846569485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,51200,0.09967733091778225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,16384,0.048042668236626514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,10240,0.04157688882615831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,512,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,12288,0.029909332593282063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,8192,0.03486311104562547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,12288,0.03690933187802633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,10240,0.025511110822359722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,10240,0.03189511100451151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,7168,0.030099554194344416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,8192,0.021856000026067097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,8192,0.02796444296836853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,6144,0.026502221822738647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,7168,0.01959288948112064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,7168,0.024122666981485155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,5120,0.02296177711751726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,6144,0.018039110634062026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,6144,0.02160444524553087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,51200,0.18649955590566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,5120,0.015986666083335876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,3584,0.016543111867374845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,65536,0.12506311469607884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,5120,0.018269333574506972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,4096,0.014614222778214348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,3584,0.013156444662147097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,4096,0.015680889288584392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,2560,0.012479110724396177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,3584,0.014442667365074158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,3072,0.012470222181744046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,2048,0.010672888822025724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,2560,0.011545778148704104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,3072,0.013208889298968844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,2560,0.01200711147652732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,1536,0.008833777573373582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,8192,0.03478755553563436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,1024,0.007112888826264276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,1536,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,2048,0.010955555571450127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,768,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,1536,0.009649777577983009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,1024,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,512,0.004233777936961916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,1024,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,768,0.007255111303594377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,4096,0.019967110620604623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,256,0.0034479999707804788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,512,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,128,0.003107555624511507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,256,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,3072,0.014663111832406787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,256,0.005975111077229182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,128,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,64,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,32,0.0028106667515304354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,64,0.006187555690606435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,65536,0.20717422167460123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,32,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,2048,0.009958222508430481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,51200,0.15561066733466253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,65536,0.16664000352223715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,16384,0.05427111188570658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,65536,0.12104533778296576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,51200,0.12953600618574354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,12288,0.04301599992646111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,768,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,10240,0.03494666682349311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,51200,0.09620622131559585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,16384,0.04615288972854614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,512,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,12288,0.02825066778394911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,8192,0.029469334416919287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,7168,0.025368889172871906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,10240,0.02401955591307746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,10240,0.03051466743151347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,128,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,6144,0.022375999225510493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,8192,0.02611911131276025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,7168,0.019015999303923715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,5120,0.019223110543357003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,7168,0.02273599969016181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,6144,0.01736177835199568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,4096,0.016454221473799813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,6144,0.020133333073721994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,5120,0.015683554940753512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,3584,0.0138568886452251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,5120,0.017607111069891188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,4096,0.014535999960369535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,3072,0.012434666355450949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,4096,0.015319999721315173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,2560,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,3072,0.012103999654452005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,3072,0.013094222380055321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,2048,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,2560,0.010438222024175854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,16384,0.036130666732788086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,2560,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,1536,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,2048,0.008980444735950893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,1024,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,12288,0.03520355621973673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,1536,0.008378666308191087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,2048,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,8192,0.02132266594303979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,768,0.004577777865860197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,1536,0.008993777963850234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,512,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,1024,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,768,0.007032000356250339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,1024,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,256,0.003160888950030009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,768,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,512,0.0069208890199661255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,128,0.0028568889117903183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,512,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,64,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,256,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,32,0.002875555513633622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,128,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,256,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,128,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,3584,0.01310488912794325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,64,0.00608711110221015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,3584,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,65536,0.17796622382269967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,32,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,16384,0.048320889472961426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,51200,0.14001154899597168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,12288,0.03741600116093954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,65536,0.12644622060987684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,65536,0.07135733630922106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,51200,0.09645955430136786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,16384,0.03525955478350321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,16384,0.023062222533755835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,12288,0.018593778212865193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,8192,0.026613333159022864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,10240,0.01641688909795549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,12288,0.027309333284695942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,7168,0.023054222265879314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,10240,0.0236835562520557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,6144,0.020023999942673575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,8192,0.020626667473051283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,7168,0.013454222016864352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,5120,0.017052445146772597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,7168,0.017846221725145977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,6144,0.016100444727473788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,4096,0.014491554763582019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,5120,0.011413333316644033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,3584,0.011695110963450538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,5120,0.01427288850148519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,4096,0.01071999967098236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,3072,0.010649777948856354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,4096,0.013088000317414602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,3584,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,2560,0.009314667019579146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,3584,0.011755555868148804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,51200,0.05793688694636027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,10240,0.03132266799608866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,3072,0.010827555424637265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,2048,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,1536,0.006719999843173557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,2560,0.009992889232105678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,2048,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,8192,0.014471999473041959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,1024,0.005365333209435145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,2048,0.009142222503821055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,1536,0.006994666324721442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,768,0.004137777619891697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,1536,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,512,0.0035004446076022256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,1024,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,768,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,1024,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,6144,0.012464000119103326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,256,0.0029395555870400537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,128,0.0028355556229750314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,768,0.006218666831652324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,256,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,512,0.005881777654091517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,64,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,128,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,32,0.0028400001012616684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,256,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,128,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,65536,0.1602124505572849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,64,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,32,0.006048000107208888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,3072,0.008612444831265343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,65536,0.12276444170210098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,51200,0.12689867284562853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,16384,0.04887911015086704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,65536,0.06864000029034086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,2560,0.008375999828179678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,12288,0.034612443712022566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,10240,0.02881066666709052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,51200,0.05595733059777153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,51200,0.09436355696784125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,8192,0.02454311152299245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,12288,0.018004443910386827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,16384,0.021956443786621094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,12288,0.026838221483760413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,7168,0.0214044451713562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,10240,0.022986667023764715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,8192,0.01425244410832723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,6144,0.018738667170206707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,7168,0.013095999757448832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,8192,0.019575110740131803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,512,0.006134222365087933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,5120,0.015850666496488783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,7168,0.0173342227935791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,6144,0.012179555164443122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,5120,0.011153777440388998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,6144,0.01586666703224182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,3584,0.011336888704035016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,5120,0.014114666316244336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,4096,0.009880888793203566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,4096,0.012799111505349478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,3072,0.010102222363154093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,3584,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,2560,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,3584,0.011694221860832639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,3072,0.008356444537639618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,2048,0.007813333637184566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,3072,0.011015999648306104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,2560,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,1536,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,2560,0.009705777797434065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,16384,0.03444088829888238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,2048,0.007251555720965068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,1024,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,2048,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,768,0.004218666503826777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,1536,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,1024,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,1536,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,512,0.0035546666218174826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,1024,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,768,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,256,0.0030462222380770575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,768,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,4096,0.013740444348918067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,512,0.006103111224042044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,128,0.0028844444702068963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,64,0.0025759999536805684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,256,0.005768888940413793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,32,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,128,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,256,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,128,0.005516444643338521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,64,0.00565155554148886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,65536,0.14143644438849554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,32,0.00620177760720253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,51200,0.11547556188371445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,65536,0.11815288331773545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,65536,0.06643199920654297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,51200,0.05372000071737501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,51200,0.08937066793441772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,16384,0.021156445145606995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,16384,0.033246222469541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,10240,0.02341777748531765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,12288,0.017455110947291057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,12288,0.025264889001846313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,8192,0.020081778367360432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,10240,0.015448888142903646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,10240,0.02200266718864441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,512,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,7168,0.01720266706413693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,8192,0.013799111048380533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,8192,0.018915555543369718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,7168,0.0127688886390792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,5120,0.012815111213260226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,7168,0.01714933415253957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,6144,0.01556799974706438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,4096,0.01146488885084788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,16384,0.04273688793182373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,5120,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,3584,0.01019200020366245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,5120,0.013777777552604675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,4096,0.00905777762333552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,10240,0.01590399940808614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,3072,0.008917333351241218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,4096,0.0123831108212471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,2560,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,3584,0.008767999708652496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,3584,0.011391999820868174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,3072,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,2048,0.007130666739410824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,3072,0.01068800025516086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,1536,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,2048,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,2560,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,1024,0.004305777864323722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,2048,0.008027555214034187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,1536,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,6144,0.01484444406297472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,768,0.004132444245947732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,1024,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,512,0.0032319999817344877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,1024,0.006568000134494569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,6144,0.011810666984981961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,768,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,256,0.0028311111446883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,768,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,512,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,128,0.002464888824356927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,512,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,256,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,64,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,256,0.005542222410440445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,128,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,32,0.0024542222834295696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,12288,0.03344000048107571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,128,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,65536,0.11521777841779922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,64,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,32,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,51200,0.09290488560994466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,2560,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,65536,0.06401422288682726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,65536,0.10993066761228774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,12288,0.029288000530666772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,51200,0.08611022101508246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,16384,0.020285333196322124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,10240,0.02587466604179806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,16384,0.031501332918802895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,1536,0.007229333122571309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,12288,0.016787555482652452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,8192,0.018301332990328472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,12288,0.024272888898849487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,7168,0.01551555593808492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,10240,0.014892444014549255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,8192,0.013397333522637686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,10240,0.0210533340771993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,6144,0.013556444810496436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,7168,0.012488889197508493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,8192,0.018616888258192275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,5120,0.011886222494973077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,6144,0.011358222199810876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,4096,0.010632889138327705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,6144,0.015126221709781222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,5120,0.010129777921570672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,3584,0.008966222405433655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,5120,0.013427555561065674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,4096,0.00925155563486947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,3072,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,4096,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,3584,0.008036444584528605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,2560,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,16384,0.03754222061898973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,3072,0.007848888635635376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,2048,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,51200,0.05189333359400431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,3072,0.009826666778988307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,2560,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,2048,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,1024,0.004212444441186057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,2048,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,1536,0.006695999867386288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,768,0.003798222376240624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,1536,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,1024,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,512,0.0031831111345026228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,768,0.006247111078765657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,7168,0.016523554921150208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,256,0.002847111059559716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,768,0.005967111223273807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,512,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,128,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,512,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,256,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,64,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,256,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,128,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,32,0.002504000026318762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,128,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,3584,0.011131555669837527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,65536,0.09142933289210002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,32,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,51200,0.07472622394561768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,2560,0.008558221989207799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,65536,0.06829955842759874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,16384,0.03032711148262024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,51200,0.04981955554750231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,51200,0.054188443554772266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,12288,0.023981332778930664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,16384,0.019693333241674636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,10240,0.02071644365787506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,16384,0.020806221498383414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,1024,0.006537777682145436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,12288,0.0163093325164583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,12288,0.016531555189026725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,10240,0.014595554934607612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,7168,0.013387555877367655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,10240,0.014762666490342883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,8192,0.012804444465372296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,8192,0.013275555438465543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,6144,0.011760888828171624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,7168,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,5120,0.01035911093155543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,6144,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,7168,0.011966221862369113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,64,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,4096,0.009520000053776635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,6144,0.011310222248236338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,3584,0.00816177742348777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,5120,0.009383111364311641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,4096,0.008452444440788692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,5120,0.010123555858929953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,3072,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,65536,0.06173955731921726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,3584,0.00795022232664956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,2560,0.006555555595291986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,3584,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,3072,0.007814221911960179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,2048,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,3072,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,2560,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,1536,0.004690666579537922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,8192,0.014907555447684394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,2048,0.007132444116804335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,1024,0.0035342222286595237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,2048,0.006688000013430913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,1536,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,768,0.0032151111712058387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,1536,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,1024,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,512,0.002862222285734283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,1024,0.0057813334796163775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,768,0.006223111102978389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,256,0.002475555572244856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,768,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,512,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,1536,0.00519466648499171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,512,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,256,0.005872000008821487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,64,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,256,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,4096,0.009189333352777693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,32,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,128,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,65536,0.0880168875058492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,64,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,32,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,2560,0.0069208890199661255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,51200,0.06981866889529757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,65536,0.06060622135798136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,16384,0.027302222119437322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,65536,0.066685332192315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,12288,0.02165511084927453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,51200,0.04863111178080241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,51200,0.052469333012898765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,16384,0.019524445136388142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,10240,0.019659555620617337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,16384,0.020132443971104093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,12288,0.01612355477280087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,8192,0.014716444744004143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,10240,0.014460444450378418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,12288,0.015951111912727356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,7168,0.012883555557992725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,10240,0.014172444740931192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,8192,0.012443555725945367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,128,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,6144,0.011360888679822287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,8192,0.013175999952687157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,7168,0.011401777466138205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,7168,0.011884444289737277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,4096,0.009073778159088558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,5120,0.009119110802809397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,6144,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,3584,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,5120,0.009655999640623728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,4096,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,3072,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,4096,0.00850755555762185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,3584,0.008074666890833113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,2560,0.0064275554484791225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,3584,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,3072,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,2560,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,3072,0.007540444533030192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,1536,0.004399111287461387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,2048,0.007204444044166141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,2048,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,1024,0.0035031110876136353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,1536,0.006903999795516332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,1536,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,768,0.003208000005947219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,5120,0.010000000397364298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,1024,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,512,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,1024,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,6144,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,256,0.002672000063790215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,768,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,512,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,128,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,512,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,256,0.005741333381997214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,64,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,256,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,128,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,32,0.0022044444663657085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,128,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,2048,0.005490666462315454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,65536,0.07742489046520658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,64,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,51200,0.062459554937150746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,2560,0.006851555572615729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,32,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,65536,0.05952178107367622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,16384,0.023455111516846552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,65536,0.06430755721198188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,51200,0.04743644595146179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,12288,0.019561777512232464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,51200,0.05056888858477274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,16384,0.019152000546455383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,10240,0.016593777471118502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,16384,0.01977422171168857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,12288,0.015871110889646742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,8192,0.014377777775128683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,10240,0.014144000079896716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,128,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,7168,0.012825777961148156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,768,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,10240,0.013943110903104147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,8192,0.011707555916574268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,6144,0.011528888510333167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,5120,0.00963466697269016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,7168,0.01072799993885888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,8192,0.012568000290128918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,6144,0.009721777505344814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,7168,0.011412444214026133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,5120,0.009233777721722921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,3584,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,4096,0.00833511104186376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,5120,0.009729777773221334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,3072,0.006907555378145642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,4096,0.008255110846625434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,3584,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,2560,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,3584,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,3072,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,2048,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,3072,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,2560,0.007674666742483775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,1536,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,2560,0.00666933341158761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,2048,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,1024,0.003839111162556542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,2048,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,12288,0.01574577722284529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,1536,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,768,0.003504000190231535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,1536,0.006056889063782162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,512,0.0032231110251612137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,1024,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,1024,0.005664000080691443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,256,0.0028124445428450904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,4096,0.008662222160233391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,768,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,128,0.002540444541308615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,6144,0.010584000084135266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,512,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,64,0.002193777718477779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,256,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,32,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,256,0.005016889009210798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,128,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,65536,0.06602044237984551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,64,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,51200,0.05412266651789347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,65536,0.05826222234302097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,65536,0.0611048870616489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,16384,0.021380444367726643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,51200,0.04639822244644165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,12288,0.0174666659699546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,51200,0.04859288864665561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,16384,0.018979556030697294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,10240,0.015045333239767285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,16384,0.018954666124449838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,12288,0.01553244392077128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,8192,0.013439999686347114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,12288,0.015446222490734525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,512,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,10240,0.013070222404268054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,7168,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,8192,0.011427555647161273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,10240,0.013482666677898832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,6144,0.010679110884666443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,8192,0.012533333566453723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,7168,0.01035377797153261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,128,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,7168,0.011239111423492432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,6144,0.009824000298976898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,4096,0.008017777568764156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,6144,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,5120,0.008973333570692275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,3584,0.007527111305130853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,5120,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,32,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,4096,0.008292444050312042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,3072,0.006806222101052602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,4096,0.00790666706032223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,2560,0.0058471109304163195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,3584,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,3072,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,2048,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,3584,0.007596444752481248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,2560,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,3072,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,1536,0.004187555362780889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,2560,0.006675555474228329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,2048,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,1024,0.003552000141806073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,2048,0.006302222195598815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,1536,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,768,0.003176888864901331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,768,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,1536,0.006142222219043308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,1024,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,1024,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,768,0.006008000009589725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,256,0.0025315555847353404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,768,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,128,0.0021484444538752237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,512,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,256,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,64,0.0021102221475707158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,256,0.004959110998445087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,32,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,128,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,128,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,65536,0.05456711186303032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,64,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,32,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,51200,0.04410755634307861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,65536,0.05678577555550469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,65536,0.05869688590367635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,16384,0.016880000631014507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,51200,0.045124444696638316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,12288,0.013800000150998434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,51200,0.04670222269164192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,16384,0.01852799952030182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,512,0.002875555513633622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,10240,0.012062222593360476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,12288,0.014721777704026965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,8192,0.011936888926559024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,12288,0.014877332581414117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,512,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,10240,0.012378666963842181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,7168,0.010431999961535135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,5120,0.009333333207501305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,8192,0.010946667028797997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,6144,0.009720889230569204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,7168,0.010405333505736457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,8192,0.011761777930789523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,5120,0.009112888740168678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,7168,0.010606221854686737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,6144,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,4096,0.007939555578761632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,5120,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,6144,0.009311999711725447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,3584,0.007232888705200619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,5120,0.008738666772842407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,4096,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,3072,0.006326222171386083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,4096,0.007962666451931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,3584,0.008039111064540016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,2560,0.00544088871942626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,3072,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,3584,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,2048,0.004554666578769684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,3072,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,2560,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,1536,0.003938666648334927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,16384,0.018539556198649935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,1024,0.0033333332588275275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,2560,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,768,0.002879111096262932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,2048,0.0063031112982167145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,1024,0.006309333360857434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,1536,0.005967111223273807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,512,0.002609777781698439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,768,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,1024,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,256,0.002295111202531391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,10240,0.013407111167907715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,768,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,128,0.002116444417172008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,512,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,64,0.0020417778028382198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,256,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,512,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,32,0.0018844445132546956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,128,0.005843555761708154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,65536,0.0414266652531094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,128,0.004951111144489712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,51200,0.03330399923854404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,32,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,65536,0.05568355321884155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,65536,0.05648710992601183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,16384,0.01348444405529234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,51200,0.04447911183039347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,51200,0.04495288928349813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,12288,0.012445333103338877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,16384,0.01679111189312405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,10240,0.011032000184059143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,16384,0.018030222919252183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,12288,0.013432888521088494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,1536,0.006599111275540457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,8192,0.009871111147933537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,12288,0.014766222900814481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,10240,0.012103999654452005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,7168,0.009319110876984065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,10240,0.012889777620633444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,8192,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,6144,0.008736888567606607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,8192,0.010999110837777456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,7168,0.010280888941552904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,5120,0.007930667036109501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,7168,0.010000000397364298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,6144,0.009532444179058075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,4096,0.007605333295133378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,6144,0.009365333451165093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,64,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,5120,0.008977777428097194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,5120,0.008630221916569604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,4096,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,3072,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,4096,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,3584,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,2560,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,3584,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,3072,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,2048,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,3072,0.0070266665683852295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,2560,0.007270221908887227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,1536,0.003956444561481476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,2560,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,2048,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,1024,0.003544888976547453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,2048,0.006565333240562015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,1536,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,768,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,1536,0.006193777753247156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,1024,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,512,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,1024,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,768,0.006060444646411472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,256,0.0024497778051429326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,256,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,768,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,128,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,512,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,64,0.002181333386235767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,256,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,3584,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,32,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,128,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,256,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,65536,0.03048444456524319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,128,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,51200,0.024915556112925213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,64,0.004990222139490975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,65536,0.054726223150889076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,16384,0.011672000090281168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,32,0.005064888960785336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,12288,0.012738666600651212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,65536,0.05379288726382785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,51200,0.04418577750523885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,10240,0.011401777466138205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,51200,0.042857779396904834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,12288,0.013467555244763693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,16384,0.017518222332000732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,8192,0.010151999692122141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,12288,0.013764444324705334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,10240,0.012138666378127204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,7168,0.009476444787449306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,10240,0.0120319997270902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,8192,0.010737777584128909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,6144,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,7168,0.010326222413116032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,8192,0.010636444720957013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,5120,0.008006222546100616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,7168,0.009993777506881291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,6144,0.009375111096435124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,4096,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,5120,0.008977777428097194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,6144,0.009324444664849175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,3584,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,4096,0.008292444050312042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,5120,0.008649778034951951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,512,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,4096,0.007726222276687622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,2560,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,3072,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,3584,0.0073315559162033936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,2048,0.004516444272465176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,3072,0.007095110913117726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,2560,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,2048,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,1536,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,2560,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,2048,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,1024,0.003152888889114062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,2048,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,768,0.002875555513633622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,1536,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,1536,0.006089777582221561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,1024,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,512,0.0025111111915773815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,1024,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,768,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,768,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,256,0.0022044444663657085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,512,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,128,0.0021520000365045336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,512,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,256,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,64,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,256,0.00508355556262864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,3072,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,32,0.0019342222561438878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,128,0.005630222045713001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,65536,0.02545333405335744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,3584,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,51200,0.020604444874657523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,64,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,65536,0.054401778512530856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,32,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,16384,0.012456888953844706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,65536,0.052892443206575185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,51200,0.04359377755059136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,12288,0.010640888578361936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,51200,0.042190220620897084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,16384,0.015963556038008798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,10240,0.009778666827413771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,16384,0.016501333978441026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,12288,0.013441777891582914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,8192,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,12288,0.013382222089502545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,10240,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,7168,0.008358221915033128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,10240,0.011939555406570435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,8192,0.010732444624106089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,6144,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,7168,0.01019377758105596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,8192,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,5120,0.007136888802051544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,7168,0.009942221972677443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,4096,0.006810666786299811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,6144,0.009172444542249044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,5120,0.008777778181764815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,3584,0.006565333240562015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,128,0.00498488876554701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,4096,0.00829777783817715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,3072,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,4096,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,3584,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,3584,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,2560,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,3072,0.0075937774446275495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,2048,0.004512888689835866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,3072,0.007172444628344641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,16384,0.016164445214801364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,1536,0.0038693332009845306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,2560,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,2560,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,2048,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,1024,0.0032133333798911837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,1536,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,2048,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,768,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,1536,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,1024,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,512,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,256,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,1024,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,512,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,6144,0.00942311104800966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,256,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,512,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,64,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,256,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,32,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,128,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,5120,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,65536,0.018890667292806838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,128,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,51200,0.015390222271283468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,64,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,65536,0.05401600069469876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,32,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,16384,0.010431110858917236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,65536,0.05196711089875963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,51200,0.04107822312249078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,12288,0.008623999853928884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,51200,0.041710221105151706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,10240,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,16384,0.016162667009565566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,12288,0.013470222552617392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,12288,0.013443555268976422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,8192,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,10240,0.01219555570019616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,7168,0.006996444529957241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,10240,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,768,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,8192,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,768,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,8192,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,7168,0.010149333212110732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,5120,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,7168,0.00976533359951443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,6144,0.009341333475377824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,4096,0.006664000037643645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,6144,0.009148444566461775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,5120,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,3584,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,5120,0.008330666356616551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,4096,0.008295999632941352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,4096,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,3584,0.00775022225247489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,3584,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,16384,0.016142222616407607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,2560,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,3072,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,2048,0.00453599997692638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,3072,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,2560,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,1536,0.0038942222793896994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,2048,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,2048,0.00630666646692488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,1024,0.003177777760558658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,1536,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,6144,0.006581333362393909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,768,128,0.0021191111041439907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,768,0.0028764444092909494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,1536,0.006048000107208888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,512,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,1024,0.006078222145636876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,1024,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,256,0.002199999988079071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,512,0.005646222167544895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,768,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,512,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,128,0.0021262222694026097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,256,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,3072,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,256,0.00519822206762102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,128,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,512,32,0.001968888979819086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,65536,0.015192889504962497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,128,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,51200,0.013159111142158508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,64,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,65536,0.047091556919945605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,32,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,16384,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,65536,0.0509111119641198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,12288,0.013065777719020844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,2560,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,51200,0.0399875541528066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,16384,0.01611911091539595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,10240,0.010104888843165504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,16384,0.016072000066439312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,8192,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,12288,0.013316444224781461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,12288,0.013410666750537025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,10240,0.012307555311255984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,7168,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,10240,0.011703111231327057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,8192,0.010690666735172272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,768,0.005882666756709416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,6144,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,5120,0.0069066666894488865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,8192,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,7168,0.010073777702119615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,4096,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,7168,0.009704888694816166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,3584,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,5120,0.008661333057615492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,6144,0.008976000050703684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,4096,0.008291555775536431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,5120,0.008272888759771982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,4096,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,3584,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,2560,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,3584,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,3072,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,2048,0.0042035554846127825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,2560,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,3072,0.0069057775868309875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,1536,0.003649777836269803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,2560,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,2048,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,1024,0.003014222201373842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,51200,0.038035555018319026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,1536,0.006430222342411677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,2048,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,1536,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,1024,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,512,0.0024746666765875286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,768,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,1024,0.005703111075692707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,256,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,768,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,256,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,512,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,128,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,256,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,64,0.0018168888572189545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,6144,0.009343110852771336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,32,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,128,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,65536,0.013878222141000958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,3072,0.005385777602593104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,128,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,51200,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,64,0.004960888789759742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,32,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,65536,0.04656444324387444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,16384,0.01034311122364468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,65536,0.04809244473775228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,51200,0.037612444824642606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,12288,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,51200,0.038733333349227905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,16384,0.015850666496488783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,10240,0.006302222195598815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,16384,0.015846222639083862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,256,768,0.0026515554636716843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,12288,0.013404444687896304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,8192,0.006794666664467917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,10240,0.012048888537618848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,12288,0.013454222016864352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,8192,0.01071999967098236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,6144,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,512,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,7168,0.010176888770527309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,5120,0.0063182223174307085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,7168,0.009647111097971598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,6144,0.009473777479595607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,4096,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,6144,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,5120,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,5120,0.008362666600280339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,3584,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,4096,0.008271999657154083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,3072,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,4096,0.007901333272457123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,3584,0.007851555943489075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,2560,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,3584,0.007332444190979004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,2048,0.004209777961174647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,3072,0.0069022224181228215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,2560,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,1536,0.003539555602603488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,2048,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,10240,0.011626666618718041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,1024,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,1536,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,8192,0.010390222072601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,768,0.0028124445428450904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,1024,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,1024,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,512,0.002513777878549364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,768,0.006042666733264923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,768,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,256,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,512,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,512,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,128,0.0021235555824306277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,256,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,3072,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,64,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,32,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,2560,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,65536,0.016703999704784818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,128,0.004846222284767362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,51200,0.014478221535682678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,2048,0.006333333336644703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,64,0.005042666776312722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,32,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,12288,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,65536,0.04794844322734409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,10240,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,51200,0.038577778471840754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,16384,0.016166221764352586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,8192,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,12288,0.01312888910373052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,7168,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,6144,0.005879111174080107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,10240,0.011371555427710215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,5120,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,8192,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,7168,0.009654222263230218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,4096,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,6144,0.008821333448092142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,5120,0.008064000142945183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,3072,0.005048000150256687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,4096,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,128,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,2560,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,2048,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,128,7168,0.0063999998900625445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,3072,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,2560,0.006562666760550604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,2048,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,1024,0.0028488888508743713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,1536,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,768,0.0025600000388092464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,1536,0.005893333504597346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,512,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,1024,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,256,0.002191999927163124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,768,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,128,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,3584,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,64,0.0018097776919603348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,256,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,32,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,128,0.004605333424276776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,65536,0.016918222109476726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,3584,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,64,0.005012444323963589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,32,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,16384,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,16384,0.005636444522274866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,65536,0.047961778110928006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,12288,0.006157333238257303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,10240,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,51200,0.03884799944029914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,8192,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,16384,0.015839111473825242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,7168,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,12288,0.013003555436929068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,10240,0.011726222104496427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,8192,0.010405333505736457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,5120,0.006471111128727595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,7168,0.009705777797434065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,4096,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,512,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,6144,0.00906844437122345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,3072,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,5120,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,2560,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,4096,0.007626666790909237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,2048,0.004552000098758274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,3584,0.007078222102589077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,51200,0.014468444718254937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,1536,0.0035093331502543557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,3072,0.006986666884687211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,1024,0.0028391112056043413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,768,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,64,1536,0.003491555651028951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,512,0.002513777878549364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,2048,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,256,0.002054222135080232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,1536,0.005653333332803514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,1024,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,128,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,64,0.0017768889665603638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,768,0.005326222214433882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,32,0.00179822225537565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,512,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,6144,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,128,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,64,0.0048071112897661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,32,0.004961777892377642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,113,32,3584,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,16384,0.3742702272203233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,16384,0.20404444800482857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,2560,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,51200,0.6176400184631348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,51200,1.1548017925686307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,12288,0.2774835427602132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,256,0.004971555537647671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,12288,0.15065599812401667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,16384,0.13418132729000515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,10240,0.236845334370931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,12288,0.10122311115264893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,51200,0.40208088027106387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,10240,0.12754844294653997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,8192,0.19263733757866752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,10240,0.0844951139556037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,8192,0.11099821991390652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,8192,0.0693111088540819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,7168,0.17029333114624023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,7168,0.09532622496287028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,10240,0.18524711661868623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,6144,0.14526222811804876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,8192,0.1500506665971544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,16384,0.29415734608968097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,7168,0.06120710902743869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,12288,0.2234640121459961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,5120,0.12195911672380234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,6144,0.07945866717232598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,6144,0.05385333299636841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,5120,0.06894399722417195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,7168,0.13543555471632215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,4096,0.10012622012032403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,5120,0.04555822081036038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,3584,0.08767910798390706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,4096,0.05534133315086365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,4096,0.04002933369742499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,3072,0.07704977856742011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,3584,0.04935555656750997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,5120,0.09826844268374973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,3584,0.034993777672449745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,4096,0.0805164443122016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,2560,0.06378755304548475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,3072,0.043154666821161904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,3584,0.07220533159044054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,2560,0.037449777126312256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,3072,0.032050665881898664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,2048,0.05187466740608215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,2560,0.02757511039574941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,2048,0.032580445210138954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,2048,0.02362222141689724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,1536,0.040463109811147056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,6144,0.1161679956648085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,2048,0.05005066593488058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,1536,0.025779555241266888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,1024,0.028350220786200628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,1536,0.01940533353222741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,1536,0.041119111908806696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,1024,0.015991111596425373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,768,0.021670222282409668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,1024,0.02058844433890449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,3072,0.06431288851632012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,1024,0.03501066565513611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,512,0.01494488947921329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,768,0.017482666505707633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,768,0.029543111721674602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,512,0.012808000048001608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,256,0.009742221898502773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,512,0.013744889034165276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,2560,0.05767822265625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,128,0.007264888948864407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,51200,0.9025973214043511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,512,0.025303999582926433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,256,0.011457777685589261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,64,0.004991111242108875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,256,0.012426666915416718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,128,0.011386666860845355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,32,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,128,0.011357333097192975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,256,0.02071377800570594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,64,0.011400889191362592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,128,0.021966222259733412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,32,0.015164444843928019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,768,0.014151111245155334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,65536,0.6336533228556315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,51200,0.4854479895697699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,65536,1.106901380750868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,51200,0.8674844635857476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,16384,0.2808328999413384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,16384,0.16612533728281656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,65536,0.4026942253112793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,51200,0.31520266003078884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,12288,0.12367822064293756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,16384,0.10861421955956353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,10240,0.16058044963412815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,12288,0.08483822478188409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,10240,0.1065244476000468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,8192,0.1340328852335612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,10240,0.0675519969728258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,12288,0.21170577738020155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,8192,0.08622133069568211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,16384,0.23881777127583823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,12288,0.17897510528564453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,7168,0.11478400230407715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,8192,0.05607555309931437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,10240,0.15117511484358045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,6144,0.09979910982979669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,7168,0.07477244403627184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,7168,0.05121511220932007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,8192,0.12196977933247884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,5120,0.08561955557929145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,6144,0.06460266643100314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,6144,0.0432640016078949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,5120,0.054042667150497437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,5120,0.03736089004410638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,7168,0.10769689083099365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,6144,0.09529955519570245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,4096,0.032227555910746254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,3584,0.06155822012159559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,4096,0.04468889037768046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,5120,0.08181244134902954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,3072,0.05407466822200351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,4096,0.06893777847290039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,3584,0.028872887293497723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,3584,0.04023911224471198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,2560,0.04575644599066841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,3072,0.03526488939921061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,3072,0.026143999563323125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,2560,0.030841777722040813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,2560,0.02295911146534814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,2048,0.0379688905345069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,3072,0.05598755677541097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,4096,0.07151289118660821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,51200,0.7330488628811307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,1536,0.029409776131312054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,2048,0.027122666438420612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,2048,0.01938311093383365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,2560,0.04943111207750109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,1024,0.02199288871553209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,2048,0.04346222347683377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,1536,0.01661066710948944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,1024,0.017465778523021273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,1024,0.013743110828929476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,768,0.01642933322323693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,1536,0.03614844547377692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,768,0.012461333639091916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,1024,0.029927998781204224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,512,0.011407111254003314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,768,0.014490667316648694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,3584,0.06334577666388617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,512,0.01142133358452055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,768,0.02602044410175747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,256,0.008251555263996124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,512,0.011973333027627734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,128,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,256,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,256,0.0107342220014996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,512,0.022015111313925848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,64,0.00452622233165635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,128,0.00981244444847107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,256,0.018777777751286823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,32,0.0045422220395671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,64,0.00998222248421775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,128,0.014264000786675347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,1536,0.02132266594303979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,32,0.012988444831636218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,65536,0.33194043901231557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,51200,0.26186754968431264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,65536,0.22752888997395834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,16384,0.08628888924916585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,51200,0.17567911412980822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,51200,0.11351466178894043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,128,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,12288,0.06646400027804904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,16384,0.06212977568308512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,16384,0.03994222150908576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,12288,0.046406222714318164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,12288,0.03186488813824124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,10240,0.05557511250178019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,51200,0.16476178169250488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,12288,0.04740177922778659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,10240,0.03972711165746053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,65536,0.96396181318495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,65536,0.14176355467902288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,65536,0.20720267295837402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,10240,0.027865777413050335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,7168,0.03959200117323134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,10240,0.04147555430730184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,8192,0.024051555328898962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,8192,0.03498488995763991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,6144,0.03451733456717597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,16384,0.06148178047604031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,7168,0.030127998855378892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,7168,0.021312000022994146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,5120,0.03015377786424425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,8192,0.03597955571280585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,6144,0.01902844508488973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,6144,0.02652888827853733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,4096,0.025771554973390367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,7168,0.033054222663243614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,5120,0.017893332574102614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,5120,0.02304711110062069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,3584,0.022230222821235657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,6144,0.029687109920713637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,4096,0.015118221441904703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,4096,0.019478221734364826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,5120,0.026594666971100703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,3584,0.014149333039919535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,3584,0.017507554756270517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,4096,0.02348711093266805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,2560,0.016255999604860943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,8192,0.04694844285647074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,3072,0.013103110922707452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,2048,0.01350044459104538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,2560,0.013809777796268463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,2560,0.012072000238630505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,2048,0.01239377756913503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,2048,0.010880889164076911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,2560,0.018111111389266122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,1536,0.010759111079904767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,2048,0.016809778081046212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,1024,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,1536,0.009524444739023844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,3072,0.01921777758333418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,1536,0.015446222490734525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,768,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,1024,0.009520889156394536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,1024,0.008064000142945183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,3072,0.015828443898095023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,512,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,3584,0.0218248897128635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,768,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,768,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,256,0.0035386664999855887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,1024,0.013995556367768182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,3072,0.01943111088540819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,512,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,128,0.0031822222388452957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,768,0.012296888563368054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,512,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,256,0.006703111032644908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,1536,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,64,0.0028311111446883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,256,0.006568888823191325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,512,0.010741333166758219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,32,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,256,0.009741333623727163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,128,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,128,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,64,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,128,0.00832088871134652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,65536,0.2420346736907959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,65536,0.17692977852291533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,65536,0.1250808901256985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,51200,0.13503200478023952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,16384,0.06485155555937025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,51200,0.10072088903850979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,16384,0.04835377799140083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,12288,0.04898133211665683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,16384,0.037018666664759316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,16384,0.05275466707017687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,12288,0.03676355547375149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,12288,0.031168000565634832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,51200,0.13599022229512533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,32,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,65536,0.1707271072599623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,8192,0.034741332133611046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,12288,0.042203555504480995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,51200,0.18704711066351998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,10240,0.025791999366548326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,10240,0.031586666901906334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,7168,0.03009866674741109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,8192,0.022759111391173467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,8192,0.028243554963005915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,6144,0.026383111874262493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,10240,0.036919110351138644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,7168,0.019673777951134574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,7168,0.024308444725142583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,5120,0.02296177711751726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,8192,0.032681776417626276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,6144,0.017806222041447956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,6144,0.021661332911915247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,4096,0.019488000207477145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,7168,0.029672000143263075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,5120,0.016833777228991192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,5120,0.018800000349680584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,3584,0.016716443830066256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,6144,0.027092445227834914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,4096,0.014551111393504672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,3072,0.0144195556640625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,4096,0.015680889288584392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,5120,0.0236462222205268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,3584,0.013383111192120446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,2560,0.012444444828563266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,3584,0.014418666561444601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,4096,0.021011556188265484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,10240,0.04125244418780009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,2048,0.01056177748574151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,3072,0.012780444489585029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,3072,0.013421333498424955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,2560,0.01144977741771274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,1536,0.008736888567606607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,2560,0.012044444680213928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,3072,0.01870133313867781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,2048,0.009945777555306753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,1024,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,2048,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,1536,0.008979555633332994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,1536,0.009581333233250512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,768,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,1024,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,1024,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,1536,0.014924445086055331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,512,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,1024,0.013430222041077085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,768,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,768,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,256,0.0038124442928367188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,512,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,512,0.006647999915811751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,768,0.011539555258221097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,128,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,256,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,256,0.006090666684839461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,512,0.011001778145631155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,3584,0.01982399986849891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,64,0.0028435554769304064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,32,0.003096888876623578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,128,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,128,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,256,0.009855111440022787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,64,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,128,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,65536,0.20689333809746635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,2560,0.017311111092567444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,32,0.0070879997478591064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,2048,0.015840889679061044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,16384,0.05422755413585239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,51200,0.15669777658250597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,65536,0.166959113544888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,51200,0.09711377488242255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,65536,0.1212879949145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,51200,0.12950755490197077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,12288,0.04239733351601494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,16384,0.04621511035495334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,16384,0.03544800149069892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,10240,0.03510844376352098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,12288,0.02772533396879832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,12288,0.03512622250450982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,16384,0.04998222324583265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,8192,0.02960977620548672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,10240,0.02442577812406752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,12288,0.04040088918473985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,51200,0.12668800354003906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,65536,0.1586515506108602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,7168,0.025576889514923096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,8192,0.026603556341595117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,10240,0.03548355566130744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,6144,0.022164444128672283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,7168,0.019328888919618394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,7168,0.022804444034894306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,8192,0.0307279990779029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,5120,0.019285332825448778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,6144,0.020298666424221463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,7168,0.02856711215443081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,6144,0.017234666479958426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,4096,0.016553777787420485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,5120,0.015607999430762397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,5120,0.01753777762254079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,6144,0.0259226659933726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,4096,0.015272888872358533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,4096,0.014096889230940076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,5120,0.02295822236273024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,10240,0.030518223841985066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,3584,0.013120888835854001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,4096,0.020628444022602506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,3584,0.014147554834683737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,8192,0.020794666475719877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,3072,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,2560,0.010682666467295753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,3584,0.019312888383865356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,3072,0.013005333642164866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,2048,0.00924444446961085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,2560,0.010570666856235929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,3072,0.017988445030318368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,2560,0.011714666253990598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,1536,0.00758844448460473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,2048,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,2048,0.009469333622190688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,2560,0.016875555117925007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,1536,0.008450667063395182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,2048,0.0158924443854226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,1536,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,768,0.004591111093759537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,1024,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,1024,0.00758133331934611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,1536,0.014428445034556918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,512,0.0037964445849259696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,3072,0.012271111210187277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,1024,0.012136889000733694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,768,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,256,0.003192000091075897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,512,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,768,0.010832000109884473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,256,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,256,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,128,0.0028364445186323593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,512,0.011052444577217102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,128,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,256,0.009424000150627559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,1024,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,64,0.002554666664865282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,128,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,32,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,128,0.008093333078755273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,3584,0.013864888913101621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,65536,0.17990133497450086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,64,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,32,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,65536,0.1261048846774631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,51200,0.1400079992082384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,768,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,65536,0.07228355275260077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,16384,0.047423998514811196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,51200,0.09667199850082397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,512,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,51200,0.0579964452319675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,16384,0.03578844335344102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,12288,0.027130666706297133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,10240,0.031071109904183283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,51200,0.11349956194559734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,65536,0.1417644421259562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,10240,0.02380888826317257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,10240,0.01643200053109063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,8192,0.02646488944689433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,10240,0.03342133429315355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,8192,0.020419556233617995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,8192,0.014797333214018079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,8192,0.02931200133429633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,12288,0.037703110112084284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,16384,0.023289778166347083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,7168,0.017973333597183228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,7168,0.013463111387358772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,6144,0.019880889190567862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,16384,0.04592622319857279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,7168,0.026325333449575637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,12288,0.037667555941475764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,5120,0.016970665918456185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,6144,0.012457778056462606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,6144,0.016206221448050607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,4096,0.014752000570297241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,5120,0.011487111449241638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,5120,0.014316444595654806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,6144,0.02442933287885454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,3584,0.013048888908492194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,4096,0.012749333348539142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,4096,0.010606221854686737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,7168,0.023028444912698533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,3072,0.01168444421556261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,3584,0.009368889033794403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,4096,0.02000266644689772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,3072,0.008795555267069075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,3584,0.01868266695075565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,3072,0.011043555206722684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,2560,0.010145777629481422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,12288,0.01850488947497474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,2048,0.008783111141787635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,2560,0.008369777765538957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,2560,0.010097777677906884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,2048,0.009008888569143083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,1536,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,2560,0.0163093325164583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,2048,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,1536,0.007061333292060428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,2048,0.01567911108334859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,1024,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,5120,0.022043555974960327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,1536,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,768,0.004531555705600315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,1536,0.012735111018021902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,1024,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,1024,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,512,0.0038480001191298165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,768,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,768,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,1024,0.011413333316644033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,256,0.0031928889867332247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,512,0.006070222291681502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,512,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,128,0.0028248888750871024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,768,0.010675555302037133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,3072,0.017624000708262127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,256,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,64,0.0026071110947264563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,512,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,128,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,32,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,128,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,256,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,65536,0.15915200445387098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,64,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,128,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,32,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,51200,0.1273733377456665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,65536,0.1220053301917182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,3584,0.011744000017642975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,16384,0.04821955495410495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,65536,0.06877155436409844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,51200,0.09462755256228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,12288,0.034916443957222834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,16384,0.021990221407678392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,16384,0.03450489044189453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,12288,0.026495999760097925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,65536,0.1408124499850803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,51200,0.11275466283162434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,12288,0.01789955629242791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,16384,0.04568799999025133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,256,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,8192,0.02458311120669047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,10240,0.01587022178702884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,10240,0.02274488906065623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,7168,0.02142488956451416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,8192,0.020071110791630216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,8192,0.0143857780430052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,10240,0.032736887534459434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,6144,0.01872977779971229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,7168,0.01309866706530253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,8192,0.028554668029149372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,7168,0.017442666822009616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,5120,0.015927111109097798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,6144,0.012135999898115793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,6144,0.015962666935390897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,7168,0.025901334153281316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,4096,0.013811555173661975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,5120,0.011061333119869232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,5120,0.013952889376216464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,6144,0.023522666758961145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,3584,0.01110666659143236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,4096,0.012821333275900947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,10240,0.028774221738179524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,5120,0.021583999196688335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,12288,0.037215110328462385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,3584,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,3072,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,4096,0.0195795562532213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,3584,0.011750222080283694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,2560,0.008637333081828224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,3072,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,3584,0.018417777286635507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,3072,0.010840000377760993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,2048,0.007685333490371704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,2560,0.00983199973901113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,3072,0.01719911065366533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,51200,0.05577066871854994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,2048,0.00830844458606508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,2048,0.007256000406212277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,2560,0.016176000237464905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,2048,0.014747555057207743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,1536,0.006989333364698622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,1024,0.004851555658711327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,1536,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,4096,0.00980533328321245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,768,0.004226666771703296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,1536,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,1024,0.00646666685740153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,1024,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,512,0.003563555578390757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,768,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,2560,0.007687999970383114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,512,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,256,0.003129777808984121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,768,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,512,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,1536,0.006471111128727595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,512,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,128,0.002826666666401757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,256,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,256,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,64,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,128,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,128,0.005741333381997214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,256,0.009277332988050249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,32,0.0028373334142896864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,64,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,128,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,32,0.006071110980378256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,768,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,1024,0.011728000309732227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,51200,0.10427288876639472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,65536,0.11591911315917969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,65536,0.06629955768585205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,16384,0.04133955637613932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,51200,0.05369333426157633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,51200,0.0900604459974501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,16384,0.020957332518365648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,16384,0.033152000771628484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,65536,0.14010666476355657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,12288,0.017455110947291057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,12288,0.025019556283950806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,10240,0.023610666394233704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,51200,0.11230577362908258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,8192,0.02020355562369029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,10240,0.01552888916598426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,12288,0.03695288962788052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,10240,0.02176711128817664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,7168,0.017097777790493436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,65536,0.13205956088172063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,8192,0.01884888940387302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,8192,0.013696889082590738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,10240,0.03198755449718899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,7168,0.012739555703269111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,7168,0.016918222109476726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,8192,0.028254220883051556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,5120,0.013063111239009433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,6144,0.015519110692871941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,6144,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,12288,0.03415200114250183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,4096,0.011815110842386881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,5120,0.013722666435771518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,5120,0.010750222537252637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,16384,0.04416977696948581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,4096,0.009316444396972656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,3584,0.009992000129487779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,5120,0.021168000168270532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,4096,0.012350222302807702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,3072,0.009000000026490953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,3584,0.008782222039169734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,4096,0.019239111079110038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,3584,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,2560,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,3072,0.010701333483060202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,3072,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,6144,0.014900444282425774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,3584,0.01789866718981001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,2048,0.006883555402358373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,2560,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,3072,0.01696711116366916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,2560,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,1536,0.0057813334796163775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,2048,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,7168,0.025263110796610515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,2560,0.015863999724388123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,1024,0.004270222038030624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,2048,0.013752889302041797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,1536,0.006731555693679386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,6144,0.023552000522613525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,768,0.003891555385457145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,1024,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,1024,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,1536,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,512,0.0031724443866146933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,768,0.006041777630647023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,768,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,1024,0.010774222513039907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,256,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,512,0.005994666781690385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,512,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,768,0.01074044406414032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,128,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,256,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,256,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,512,0.010048888623714447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,64,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,128,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,2048,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,256,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,64,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,65536,0.11458667119344075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,1536,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,32,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,51200,0.09168799718221028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,65536,0.06448800033993192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,65536,0.10865955882602268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,51200,0.051783998807271324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,51200,0.0861404471927219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,65536,0.14005422592163086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,51200,0.11200355158911811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,16384,0.03181688984235128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,128,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,12288,0.02880355715751648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,16384,0.020297777321603563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,16384,0.04472177889611986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,10240,0.02537244392765893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,12288,0.016929777132140267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,12288,0.03635111120012071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,8192,0.0184871107339859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,10240,0.020997333857748244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,10240,0.015136889285511441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,10240,0.03165955675972833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,8192,0.013356444736321768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,7168,0.015562666787041558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,8192,0.018394667241308425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,16384,0.03675377700063918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,6144,0.013438222308953604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,7168,0.012332444389661154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,8192,0.027768888407283362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,7168,0.016534222496880423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,5120,0.011968889170222811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,6144,0.01126844435930252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,6144,0.0148062225845125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,7168,0.02513422237502204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,4096,0.010417777631017896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,128,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,5120,0.009966221948464712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,5120,0.013364444176355997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,6144,0.023193778263198003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,3584,0.009215110705958473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,4096,0.008830221990744272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,4096,0.012072000238630505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,5120,0.021185777253574792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,3072,0.008290666672918532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,3584,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,3584,0.011187555889288584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,4096,0.018920888503392536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,2560,0.007248889240953658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,3584,0.017839110559887357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,3072,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,2048,0.006351111249791251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,3072,0.01667022208372752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,2560,0.008617777791288164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,2560,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,1536,0.005511111269394557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,2560,0.015418666932317944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,2048,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,2048,0.007749333149856991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,32,0.0024968888610601425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,1024,0.0042151109211974675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,1536,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,1536,0.0069982219073507525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,768,0.0038444445365005066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,2048,0.013438222308953604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,12288,0.024614221519894067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,1024,0.006303999986913469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,512,0.003216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,1536,0.012703999876976013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,768,0.006223111102978389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,256,0.002855111120475663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,512,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,768,0.010400888820489248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,3072,0.007956444389290279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,512,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,128,0.002471999989615546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,256,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,256,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,256,0.009376889301670922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,128,0.005709333138333426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,64,0.0022462221483389535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,128,0.0053102220926019884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,32,0.0025066667132907445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,65536,0.09153599871529473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,128,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,64,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,1024,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,32,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,51200,0.07414133018917508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,1024,0.01144088887506061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,65536,0.07009955247243245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,65536,0.06201066573460897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,16384,0.03074044320318434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,51200,0.05428711242145962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,512,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,51200,0.049633777803844877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,12288,0.023916444844669763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,65536,0.13984889454311794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,16384,0.0196426659822464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,16384,0.020969778299331665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,10240,0.021074665917290583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,51200,0.11196622583601211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,12288,0.016693333784739178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,16384,0.04386933313475715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,10240,0.014799111419253878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,10240,0.014499555031458536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,12288,0.03530933459599813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,7168,0.013389333254761167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,10240,0.030793779426150854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,8192,0.012834666503800286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,8192,0.013353777428468069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,6144,0.011723555624485016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,7168,0.012053333222866058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,7168,0.011534222298198275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,5120,0.010253333383136326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,8192,0.026682666606373254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,6144,0.011204444699817233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,6144,0.01069866700304879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,7168,0.024497777223587036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,4096,0.009464888936943477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,5120,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,5120,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,6144,0.023078221413824294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,3584,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,4096,0.008980444735950893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,4096,0.008662222160233391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,5120,0.020474667350451153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,3072,0.007383999725182851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,12288,0.016240888171725802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,4096,0.018411555223994784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,3584,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,2560,0.006507555643717448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,8192,0.015253333581818474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,3072,0.00775999989774492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,3072,0.007752888732486301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,3584,0.017456000049908955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,2048,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,2560,0.00738933351304796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,2560,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,3072,0.016536888149049547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,2048,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,2048,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,2560,0.014415999253590902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,1024,0.003632889025741153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,1536,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,1536,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,2048,0.013356444736321768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,768,0.003181333343187968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,1024,0.006558222075303395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,1024,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,1536,0.011767110890812345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,512,0.0028115556471877625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,768,0.005997333261701796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,768,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,1024,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,256,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,512,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,768,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,512,0.0053297777970631914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,128,0.0024862223201327855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,3584,0.00813422186507119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,256,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,64,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,512,0.009690666364298927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,32,0.002191999927163124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,128,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,128,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,256,0.009300444689061906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,65536,0.08269600073496501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,64,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,128,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,32,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,51200,0.0662720004717509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,65536,0.06602311134338379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,65536,0.06053955687416924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,16384,0.027783112393485174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,51200,0.052116443713506065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,51200,0.04867733187145657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,65536,0.13982133070627847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,12288,0.021464889248212177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,16384,0.019847111569510568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,16384,0.01943111088540819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,51200,0.11145955986446804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,10240,0.019896888070636325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,12288,0.016233777006467182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,12288,0.016068443655967712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,256,0.005107555538415909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,16384,0.04279111160172356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,8192,0.014714666538768344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,10240,0.014371555712487964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,12288,0.03513244456715054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,7168,0.01275199982855055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,1536,0.0046817776229646476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,8192,0.013056000073750814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,8192,0.012439111040698158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,6144,0.011384000380833944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,10240,0.030543999539481267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,8192,0.026512000295850966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,7168,0.011553777588738335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,5120,0.010128888818952773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,6144,0.01036977767944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,6144,0.010891555911964841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,7168,0.02475111186504364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,4096,0.008655110994974772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,5120,0.009983110758993361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,6144,0.02243999971283807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,4096,0.0086666668454806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,5120,0.020262221495310467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,4096,0.01891911029815674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,10240,0.014135110709402295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,3584,0.007679999702506595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,3072,0.007258666886223688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,3584,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,3584,0.017444445027245414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,2560,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,3072,0.0074373334646224976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,7168,0.011236444115638733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,3072,0.015771556231710646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,2560,0.007335111498832703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,2048,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,2560,0.00674933319290479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,5120,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,2048,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,2560,0.014133334159851074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,1536,0.004527111020353106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,2048,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,1024,0.0034773333205117118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,4096,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,1536,0.006700444552633498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,2048,0.01311911145846049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,1536,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,768,0.003162666741344664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,1024,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,1024,0.005839111076460944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,512,0.0028115556471877625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,1536,0.012104888757069906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,768,0.006002666635645761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,768,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,256,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,512,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,768,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,512,0.0053688887920644544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,3072,0.007861333588759104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,128,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,256,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,256,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,64,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,128,0.0058648888435628675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,128,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,32,0.0021342221233579847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,256,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,65536,0.07154488563537598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,64,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,3584,0.00794666674402025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,32,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,51200,0.058951112959120006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,65536,0.0638293359014723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,65536,0.05968177980846829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,16384,0.02363288899262746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,51200,0.050754666328430176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,51200,0.047535998953713306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,1024,0.011335999601417117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,65536,0.13940978050231934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,12288,0.01975644462638431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,16384,0.01937866707642873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,16384,0.019253333409627277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,10240,0.01683999929163191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,51200,0.11118488841586643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,512,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,12288,0.015782222151756287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,8192,0.014709333578745524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,16384,0.04288355509440104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,10240,0.014124444789356656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,12288,0.0347013341055976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,128,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,7168,0.013295110729005603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,8192,0.01219466659757826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,8192,0.012416889270146688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,7168,0.011041777829329172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,8192,0.02652799917591943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,6144,0.01237422227859497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,7168,0.01142133358452055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,7168,0.02458400030930837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,6144,0.010031999813185798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,5120,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,6144,0.010637332995732626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,6144,0.02268444498380025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,5120,0.00923644420173433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,4096,0.008454221818182204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,5120,0.009455111291673448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,12288,0.015847111741701763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,3584,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,4096,0.008295999632941352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,5120,0.02060622142420875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,10240,0.013953778478834363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,4096,0.008362666600280339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,3072,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,3584,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,3584,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,10240,0.030256887276967365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,2560,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,4096,0.017863111363516915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,3072,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,3072,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,2048,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,2560,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,3584,0.016536888149049547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,1536,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,3072,0.015142222245534262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,2048,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,2048,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,2560,0.01447111037042406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,1024,0.003887999802827835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,1536,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,1536,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,768,0.003506666670242945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,2048,0.013042666845851474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,1024,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,1536,0.012078222301271228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,1024,0.005623111294375525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,512,0.00314666661951277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,768,0.005538666827811136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,1024,0.01074488874938753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,256,0.0028088889602157804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,512,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,768,0.01070755554570092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,512,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,128,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,256,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,512,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,256,0.005127111242877113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,64,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,32,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,128,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,128,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,256,0.009648000200589498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,65536,0.062169777022467725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,128,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,64,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,32,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,51200,0.048807998498280845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,65536,0.061117331186930336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,16384,0.020639111598332722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,51200,0.048150221506754555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,51200,0.04618311259481642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,65536,0.13913867208692762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,768,0.006226666685607698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,16384,0.018911111685964797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,12288,0.016812443733215332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,51200,0.11110933621724446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,10240,0.014479999740918478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,12288,0.015371556083361307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,16384,0.04285600119166904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,10240,0.013636444177892474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,12288,0.034450666771994695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,10240,0.013439999686347114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,8192,0.012688889271683164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,2560,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,7168,0.011370666325092316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,10240,0.030214223596784804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,8192,0.011419555379284753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,8192,0.012092444631788464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,65536,0.05819822020000882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,6144,0.010809777511490716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,7168,0.010650667051474253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,8192,0.026192888617515564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,7168,0.011085333095656501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,5120,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,6144,0.009941333697901832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,6144,0.010041777458455827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,7168,0.024128889044125874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,4096,0.00846400029129452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,16384,0.019003555178642273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,5120,0.009186666872766282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,5120,0.008942222429646386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,6144,0.02243288854757945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,3584,0.0075804442167282104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,4096,0.00852177788813909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,5120,0.02009600069787767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,4096,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,12288,0.015575110912322998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,3072,0.006735111276308696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,3584,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,3584,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,4096,0.017824888229370117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,3072,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,3072,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,2048,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,2560,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,2560,0.006670222100284364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,3072,0.015167999598715039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,1536,0.0042275554604000514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,2048,0.0069884442620807225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,2560,0.014155555102560254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,2048,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,1024,0.00349866681628757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,1536,0.006666666517655055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,1536,0.00619911112719112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,2048,0.01328888866636488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,768,0.0031742221779293488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,1024,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,1024,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,1536,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,512,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,768,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,1024,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,2560,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,256,0.002505777817633417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,512,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,768,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,3584,0.015847999188635085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,256,0.005630222045713001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,512,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,256,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,64,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,128,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,128,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,32,0.0021271111650599376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,256,0.009464888936943477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,65536,0.05035466618008084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,64,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,128,0.007885333564546373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,32,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,51200,0.04052444299062093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,65536,0.058486223220825195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,65536,0.05674666828579373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,16384,0.016879111528396606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,768,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,51200,0.04518666532304552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,65536,0.13903733094533285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,12288,0.014124444789356656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,16384,0.018381334013409086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,16384,0.01758399936887953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,51200,0.11113156212700738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,128,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,10240,0.012186666329701742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,12288,0.014254222313563028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,16384,0.042268445094426475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,8192,0.011600889265537262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,10240,0.01333688861793942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,10240,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,12288,0.034274667501449585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,7168,0.010355555348926121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,8192,0.011186666786670685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,10240,0.03016800019476149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,7168,0.010402667025725046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,7168,0.01073688848151101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,6144,0.010332444475756751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,51200,0.046819554434882275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,6144,0.009737778041097853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,512,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,5120,0.008988444175985124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,6144,0.009662222531106737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,6144,0.021294222937689886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,4096,0.007684444387753804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,5120,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,12288,0.015125332607163323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,4096,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,5120,0.01963644391960568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,4096,0.007861333588759104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,3584,0.00720177756415473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,8192,0.012016889121797351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,3072,0.006310222049554189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,3584,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,3584,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,8192,0.026055110825432673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,2560,0.005766222046481238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,3584,0.015810666812790763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,3072,0.007631111476156447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,3072,0.007238222493065729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,2048,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,7168,0.024266666836208765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,2560,0.007345777418878343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,3072,0.015135111080275642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,2560,0.006709333509206772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,1536,0.0038480001191298165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,2048,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,2560,0.013728888498412238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,5120,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,1024,0.00319377767542998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,1536,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,1536,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,1024,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,1536,0.01205244412024816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,768,0.0029724445194005966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,1024,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,4096,0.01779911087618934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,512,0.002542222125662698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,768,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,768,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,768,0.01032977799574534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,512,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,256,0.0022044444663657085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,512,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,256,0.005199111170238919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,2048,0.0063279999627007385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,128,0.002143111079931259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,256,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,256,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,2048,0.013094222380055321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,64,0.0019120000716712738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,128,0.0058542220956749385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,32,0.002128888956374592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,128,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,65536,0.04135555691189236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,64,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,32,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,51200,0.033287998702791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,1024,0.010678222609890832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,65536,0.056313779619004994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,65536,0.05545511179500156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,16384,0.01368444412946701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,65536,0.1388444503148397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,512,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,51200,0.04474844535191854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,12288,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,16384,0.018340445227093168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,16384,0.017440888616773818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,51200,0.1106684472825792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,10240,0.011005333728260465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,12288,0.0147706667582194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,16384,0.041802667909198336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,10240,0.012770666844315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,12288,0.034323556555642024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,10240,0.01237777786122428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,8192,0.009983110758993361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,128,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,7168,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,8192,0.011071999867757162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,8192,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,6144,0.0087013335691558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,7168,0.010276444256305695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,8192,0.02586666742960612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,7168,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,51200,0.04479022158516777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,5120,0.008296888735559251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,6144,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,6144,0.009633777870072259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,7168,0.023297778434223596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,4096,0.0075902218619982404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,5120,0.008988444175985124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,12288,0.013464888764752282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,4096,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,5120,0.01885422236389584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,4096,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,4096,0.016529776983790927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,3584,0.008012444608741337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,3072,0.005727111051479976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,3584,0.007328888608349695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,10240,0.02982577681541443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,3584,0.015692444311247934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,2560,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,3072,0.01483022173245748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,2560,0.006765333314736684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,2048,0.004541333350870344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,2560,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,2560,0.013970666461520724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,5120,0.008620444271299574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,2048,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,1536,0.004182222402758068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,2048,0.006384000182151794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,6144,0.021176889538764954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,1024,0.0034933334423436057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,1536,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,2048,0.012727110750145383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,1536,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,768,0.003208000005947219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,1024,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,1536,0.011728888584507836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,3072,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,3072,0.007297777467303806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,512,0.002814222127199173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,768,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,1024,0.011019555230935415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,512,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,768,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,256,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,512,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,256,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,512,0.00980000032318963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,256,0.005116444494989184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,128,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,256,0.009137777818573846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,64,0.0021262222694026097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,32,0.0021439999755885866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,128,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,1024,0.006223111102978389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,64,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,65536,0.03080444534619649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,32,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,51200,0.024266666836208765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,768,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,65536,0.05498577819930183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,65536,0.05397955576578776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,16384,0.011765333513418833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,51200,0.04282666577233208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,51200,0.04411733481619093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,65536,0.13833778434329563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,12288,0.012533333566453723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,16384,0.01626666718059116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,16384,0.01725955473052131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,51200,0.110189331902398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,10240,0.011392888923486074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,12288,0.013416888813177744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,3584,0.006595555692911148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,12288,0.013779555757840475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,16384,0.0417520006497701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,8192,0.01016355554262797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,128,0.004955555415815777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,10240,0.012109333442317115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,12288,0.033450666401121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,7168,0.009315555294354757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,8192,0.010715555813577441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,8192,0.01035644445154402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,6144,0.009642666412724389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,10240,0.028992888000276353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,8192,0.02499733368555705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,7168,0.010061333576838175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,5120,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,6144,0.009646221995353699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,6144,0.009321778184837764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,7168,0.023166222704781428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,4096,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,5120,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,5120,0.008602666358153025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,6144,0.020644444558355544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,3584,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,4096,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,5120,0.018952889574898612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,3072,0.005643555687533484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,3584,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,3584,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,3584,0.015581332974963717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,3072,0.007080000307824876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,2560,0.005149333427349727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,10240,0.012079111403889127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,2560,0.007247111035717859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,7168,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,2048,0.004853333450025982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,2560,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,2560,0.013788444300492605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,2048,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,1536,0.0042160000238153665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,2048,0.012749333348539142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,1536,0.0060159998635451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,4096,0.007786666353543599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,1024,0.0031999999450312722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,1536,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,1536,0.011766222616036734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,4096,0.016524444023768108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,768,0.002867555452717675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,1024,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,1024,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,3072,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,512,0.002552888873550627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,768,0.005889777921968036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,768,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,1024,0.011386666860845355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,256,0.002475555572244856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,512,0.005872888697518243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,768,0.01001422190003925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,512,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,128,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,256,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,256,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,512,0.010001777774757808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,64,0.002052444343765577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,32,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,2048,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,128,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,256,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,128,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,32,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,65536,0.05305955476231045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,51200,0.020688888099458482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,65536,0.0544417765405443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,65536,0.13833244641621908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,51200,0.04219466778967115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,16384,0.012392889294359418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,3072,0.014764444695578681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,51200,0.04371910956170824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,51200,0.11011555459764268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,12288,0.010750222537252637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,16384,0.016612443659040663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,16384,0.016183111402723525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,128,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,10240,0.010016000105275048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,65536,0.02530755599339803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,12288,0.01350311107105679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,12288,0.013459555804729462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,16384,0.041654222541385226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,8192,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,64,0.005059555586841372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,10240,0.012083555261294046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,10240,0.011909333368142446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,7168,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,8192,0.011049778097205691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,12288,0.033088889386918806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,6144,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,8192,0.010406222608354356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,10240,0.02880355715751648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,7168,0.010229333407349056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,5120,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,7168,0.009977777798970541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,8192,0.025051555699772302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,6144,0.009422221945391761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,4096,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,6144,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,7168,0.022679999470710754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,5120,0.009012444151772393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,3584,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,5120,0.008572444319725037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,6144,0.02036533256371816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,4096,0.008030222521887885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,4096,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,3072,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,5120,0.018592889110247295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,3584,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,2560,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,3584,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,4096,0.016533333394262526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,3072,0.007589333587222629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,2048,0.00452888881166776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,3072,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,3584,0.016089777151743572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,2560,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,2560,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,1536,0.0038524443904558816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,3072,0.014833778142929077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,2048,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,2048,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,1024,0.0031759999692440033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,2560,0.014121777481502958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,1536,0.006311111152172089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,768,0.0028524444335036804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,2048,0.013089777694808112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,1536,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,1024,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,512,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,1024,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,1536,0.011711999773979187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,768,0.006001777946949005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,256,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,768,0.005537777725193236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,512,0.005838222387764189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,1024,0.011090666883521609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,128,0.0021182222084866627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,512,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,256,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,768,0.010399999717871347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,64,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,256,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,768,32,0.0019004444281260173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,65536,0.01921777758333418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,512,0.00998399986161126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,128,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,128,0.005074666606055366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,256,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,51200,0.015588444140222339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,64,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,32,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,128,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,65536,0.05398133397102356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,16384,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,65536,0.051656001143985324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,12288,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,51200,0.041659557157092623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,51200,0.04177422324816386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,16384,0.016145777371194627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,10240,0.008006222546100616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,16384,0.016131555040677387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,12288,0.013410666750537025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,51200,0.10991822348700629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,12288,0.013367111484209696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,16384,0.041638222005632185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,8192,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,10240,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,10240,0.011754666765530905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,7168,0.0087333329849773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,8192,0.010829333629873065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,8192,0.01036177741156684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,10240,0.02843822373284234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,6144,0.00793422261873881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,8192,0.024680889315075342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,7168,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,7168,0.009768000079525841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,5120,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,6144,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,6144,0.009215999808576372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,7168,0.023066666391160753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,4096,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,5120,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,5120,0.008754666480753157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,6144,0.020336889558368258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,3584,0.006020444548792309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,65536,0.137609773212009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,4096,0.00813066628244188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,4096,0.00790666706032223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,5120,0.01887999971707662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,3072,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,3584,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,3584,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,2560,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,12288,0.032925334241655134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,3072,0.007366221812036302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,3072,0.007104888558387756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,2048,0.0042275554604000514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,3584,0.015689777003394235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,2560,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,3072,0.015114666687117683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,1536,0.0038417776425679526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,2048,0.006903999795516332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,2048,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,2560,0.013515555196338229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,1024,0.0031582222630580268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,2048,0.013067555924256643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,1536,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,1024,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,1024,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,768,0.002855111120475663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,1536,0.011699555648697747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,768,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,512,0.0024844445288181305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,768,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,4096,0.016917333006858826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,768,0.010413332945770687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,512,0.005752888818581899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,256,0.0021537778278191886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,512,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,128,0.002151111140847206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,2560,0.00665866666369968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,256,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,512,0.010009778042634329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,256,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,64,0.001820444439848264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,128,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,128,0.005017777697907554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,256,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,1536,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,128,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,64,0.004990222139490975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,32,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,51200,0.013208889298968844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,65536,0.05177244544029236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,65536,0.04741511080000135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,16384,0.009199110998047723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,1024,0.010703999963071613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,51200,0.03797600004408095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,51200,0.04047022263209025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,16384,0.0158942225906584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,16384,0.01605688863330417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,51200,0.10934133662117852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,12288,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,10240,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,65536,0.13821421729193792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,12288,0.01311466677321328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,16384,0.04076000054677328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,12288,0.013247110777431063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,8192,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,512,32,0.0018666666001081467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,10240,0.011794666449228922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,7168,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,65536,0.015190222197108798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,8192,0.010797333386209277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,8192,0.010335110955768162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,6144,0.005988444305128521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,7168,0.010069333016872406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,7168,0.009734222458468543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,8192,0.024784889486100938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,5120,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,6144,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,6144,0.021095110310448542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,5120,0.008292444050312042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,5120,0.008782222039169734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,4096,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,12288,0.032588445478015475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,5120,0.018909333480728995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,4096,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,10240,0.011758222348160215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,3584,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,4096,0.00830844458606508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,3584,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,3072,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,3584,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,4096,0.016515556308958266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,7168,0.022608889473809138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,2560,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,3584,0.01578044394652049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,3072,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,3072,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,2048,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,2560,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,2560,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,3072,0.014489778214030795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,1536,0.003571555432346132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,2048,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,2560,0.013792888985739814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,2048,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,1024,0.002942222274012036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,1536,0.006497777584526274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,1536,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,1024,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,1024,0.006068444500366847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,1536,0.011413333316644033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,10240,0.029022223419613306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,6144,0.009331555830107795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,512,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,768,0.005347555710209741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,768,0.006074666563007567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,768,0.01034044474363327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,512,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,512,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,512,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,128,0.0019137778629859288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,256,0.005629333357016246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,256,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,128,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,64,0.001812444378932317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,256,0.009434666898515489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,2048,0.012447111308574677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,128,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,32,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,65536,0.013456000222100152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,768,0.002797333316670524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,64,0.004724444614516364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,32,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,51200,0.011696889168686338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,65536,0.048120889398786754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,1024,0.010774222513039907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,65536,0.04652977652019925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,16384,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,51200,0.037667555941475764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,65536,0.1369582282172309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,256,256,0.002238222294383579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,12288,0.006311111152172089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,16384,0.01621866722901662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,16384,0.015834665960735746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,10240,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,51200,0.10873155461417304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,12288,0.013423110875818463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,16384,0.04041688972049289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,12288,0.013455111119482251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,10240,0.012050666742854647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,10240,0.011710222396585675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,12288,0.032580445210138954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,7168,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,8192,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,8192,0.010802666346232096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,128,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,7168,0.009714666340086196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,7168,0.010056888891590966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,8192,0.025024889243973628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,6144,0.009674666656388177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,6144,0.009029332962301042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,7168,0.023048889305856492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,5120,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,6144,0.020953777763578627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,5120,0.008928889201747047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,4096,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,5120,0.008343111309740279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,8192,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,3584,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,4096,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,5120,0.01891288823551602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,4096,0.007735110819339752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,3072,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,4096,0.016917333006858826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,3584,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,10240,0.028427554501427546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,3584,0.007214222517278459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,2560,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,3072,0.007093333535724216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,6144,0.005980444451173146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,3584,0.015832889411184523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,2048,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,2560,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,2560,0.006409777949253718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,3072,0.01479022204875946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,1536,0.003530666646030214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,51200,0.03870844509866502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,2048,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,2048,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,2560,0.013812444276279874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,1024,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,1536,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,1536,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,768,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,2048,0.012401777837011548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,1024,0.006041777630647023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,1024,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,512,0.002483555633160803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,1536,0.011426666544543372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,1024,0.01107377807299296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,768,0.005435555759403441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,256,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,512,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,512,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,768,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,128,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,3072,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,512,0.009679111341635386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,256,0.005169777820507686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,64,0.0018311111877361934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,128,32,0.0018595556418100994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,128,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,65536,0.012094222009181976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,256,0.009376889301670922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,51200,0.00812533332241906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,64,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,128,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,32,0.004602666530344221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,16384,0.005893333504597346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,12288,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,10240,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,65536,0.048230220874150596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,8192,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,12288,0.013067555924256643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,128,0.007755555212497711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,7168,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,10240,0.01147555559873581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,6144,0.004958222309748332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,5120,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,16384,0.01601777805222405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,8192,0.0103493332862854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,4096,0.005324444423119227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,3584,0.00499555551343494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,6144,0.008715555899673039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,5120,0.008222222328186035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,7168,0.00980000032318963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,4096,0.007830222447713217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,2048,0.004538666870858935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,3072,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,1536,0.0035626664757728577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,3584,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,1024,0.002921777880854077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,2048,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,768,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,2560,0.006684444430801604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,1536,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,512,0.002558222247494592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,256,0.0022275555464956495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,51200,0.03871555460823907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,128,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,768,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,64,0.0018097776919603348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,1024,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,512,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,768,0.005969777703285217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,32,0.0018764444523387486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,65536,0.012055111428101858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,51200,0.008578666382365758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,128,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,16384,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,64,0.004987555659479565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,3072,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,32,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,10240,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,64,2560,0.004967111266321606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,8192,0.005498666730191972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,51200,0.03844177722930908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,7168,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,16384,0.01587288909488254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,12288,0.013075555364290873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,6144,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,5120,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,10240,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,4096,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,8192,0.010324444207880232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,3584,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,7168,0.009328000247478485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,6144,0.008808000220192803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,5120,0.00831911133395301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,4096,0.007297777467303806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,2560,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,3584,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,2048,0.004524444540341695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,3072,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,1536,0.003573333223660787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,2560,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,1024,0.0028844444702068963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,2048,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,256,0.005151999907361137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,12288,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,768,0.0025395556456512874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,1536,0.005893333504597346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,512,0.002472888885272874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,256,0.0021128888345426982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,128,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,65536,0.047873778475655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,64,0.0018604445374674266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,512,0.004969777746333016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,768,0.005658666706747479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,32,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,128,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,112,32,3072,0.004852444347408083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,64,0.0047048889100551605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,32,0.004975111120276981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,16384,0.3799351056416829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,51200,0.6087288856506348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,1024,0.005376889059940974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,16384,0.2060391108194987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,12288,0.2825395531124539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,256,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,51200,1.1487439473470051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,10240,0.20786310566796196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,12288,0.14971555603875056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,12288,0.10442488723331028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,10240,0.08428355720308091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,16384,0.13372977574666342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,51200,0.3991599877675374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,8192,0.07377955648634169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,8192,0.10796622435251872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,7168,0.09514933162265354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,7168,0.06086844205856323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,6144,0.13256977664099798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,6144,0.08101422256893583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,5120,0.10935111178292169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,4096,0.09316088755925496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,5120,0.04881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,3584,0.0791813333829244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,5120,0.06856889194912381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,4096,0.05470044414202372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,4096,0.038463112380769514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,3072,0.06876266664928861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,3584,0.034068445364634194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,3584,0.04974400003751119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,10240,0.1260897715886434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,2560,0.05792088641060723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,3072,0.03104177779621548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,3072,0.042839109897613525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,7168,0.1509759955936008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,2560,0.026708443959554035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,1536,0.03700088792377048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,2560,0.037209779024124146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,1024,0.027497778336207073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,2048,0.03197155396143595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,1536,0.01867288847764333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,1536,0.02539377742343479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,768,0.020615110794703167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,1024,0.015240000353919135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,512,0.013807999591032663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,6144,0.053146666950649686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,768,0.01385155568520228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,256,0.008586666650242275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,768,0.020614221692085266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,512,0.012483555409643384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,128,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,512,0.015405333704418607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,256,0.011392888923486074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,256,0.012472888661755456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,64,0.005196444276306364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,128,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,32,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,8192,0.17393155892690024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,128,0.011385777758227454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,2048,0.047373334566752114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,64,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,32,0.015460444821251763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,2048,0.022932445009549458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,1024,0.0255039996571011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,51200,0.4789075321621365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,51200,0.8655831019083658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,65536,0.6282702022128636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,65536,1.1087564892239041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,16384,0.16059199968973795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,12288,0.11877511607276069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,16384,0.10814933644400702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,51200,0.32033154699537486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,12288,0.21204711331261528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,65536,0.39974310663011337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,12288,0.08405511246787177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,16384,0.2807617717319065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,8192,0.15821866194407144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,10240,0.06735289096832275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,10240,0.10082311100429958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,7168,0.13873244656456843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,8192,0.060473779837290444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,6144,0.11892710791693793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,7168,0.07516000005933973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,7168,0.04933244321081373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,5120,0.10103288624021743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,6144,0.05020799901750353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,6144,0.06351821952395968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,4096,0.08149244387944539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,5120,0.03693244523472256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,5120,0.05463822351561653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,3584,0.0723208917511834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,4096,0.044481777482562594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,3072,0.06299200322892931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,3584,0.02860622273551093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,10240,0.19396532906426323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,2560,0.05295111073387993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,3584,0.03992088966899448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,3072,0.025395555628670588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,3072,0.03492710987726847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,2560,0.024746666351954143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,2048,0.04283644424544441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,1536,0.03345777922206455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,2560,0.030576000610987347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,2048,0.019369777705934312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,2048,0.026725333597924974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,1536,0.021202666891945735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,768,0.018208000395033095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,8192,0.08684533172183567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,1024,0.013177778157922955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,1024,0.021873777111371357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,512,0.01312711089849472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,768,0.016533333394262526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,256,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,768,0.012366222010718452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,4096,0.032053334845436945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,128,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,256,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,128,0.010031110710567897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,256,0.009922666682137383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,64,0.0042257776690853965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,1024,0.023973332511054143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,128,0.009857777920034196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,32,0.0046568889584806235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,1536,0.01572355628013611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,32,0.013207111093733044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,64,0.010370666782061258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,65536,0.3296497662862142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,51200,0.2607119878133138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,65536,0.227729770872328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,16384,0.08587466345893012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,512,0.01311288856797748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,65536,0.14377600616878933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,12288,0.06559022267659505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,16384,0.0407244430647956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,512,0.011189333266682096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,10240,0.054783112472958036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,51200,0.11098044448428684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,16384,0.06090044313006931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,8192,0.046925332811143666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,12288,0.045928888850741915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,7168,0.040059556563695274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,10240,0.027793778313530818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,8192,0.025797333982255723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,10240,0.039487110243903265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,6144,0.03468799922201369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,8192,0.03390222125583225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,5120,0.029667556285858154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,7168,0.021199110481474135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,7168,0.02979644470744663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,4096,0.025384889708624944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,6144,0.0265600002474255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,5120,0.016772444049517315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,3584,0.021681777305073206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,5120,0.02292977770169576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,4096,0.015269334117571512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,3072,0.018915555543369718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,4096,0.019259555472267997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,3584,0.013743110828929476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,2560,0.016009777784347534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,3072,0.01274311128589842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,3584,0.017283555534150865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,2048,0.012796444197495779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,3072,0.015572445260153877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,2560,0.01198755535814497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,1536,0.010337777435779572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,51200,0.17337600390116373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,2560,0.013968888256284924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,1024,0.00848711116446389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,2048,0.012453333371215396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,1536,0.009441778063774109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,768,0.006593777901596493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,1024,0.008012444608741337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,512,0.005061333378156026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,1024,0.011318222516112857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,768,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,768,0.008669333325492011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,256,0.0035084444615576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,512,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,6144,0.021877333521842957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,512,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,128,0.0028924445311228433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,256,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,64,0.0027395555128653846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,256,0.006568888823191325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,128,0.006732444382376141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,32,0.0028391112056043413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,128,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,64,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,32,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,65536,0.24403466118706596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,2048,0.010597333312034607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,51200,0.1869502200020684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,65536,0.17407022582160103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,12288,0.032009777095582746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,1536,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,16384,0.06318133407168917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,12288,0.049323555496003896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,65536,0.12653777334425184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,51200,0.09929244385825263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,10240,0.04200177722507053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,16384,0.04851644568973117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,12288,0.030064887470669214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,8192,0.03484088844723172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,12288,0.036369777388042875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,10240,0.0251857770813836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,10240,0.03166666626930237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,8192,0.022641777992248535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,6144,0.026696888936890498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,7168,0.01943288909064399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,7168,0.024167999625205994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,5120,0.023223110371165808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,6144,0.01753422286775377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,6144,0.021313778228229944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,4096,0.019694222344292533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,5120,0.01581333412064446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,5120,0.018239110708236694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,3584,0.0166311115026474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,4096,0.014126222994592456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,51200,0.13350489404466417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,16384,0.03717777795261807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,3072,0.014711999230914645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,3584,0.013096888860066732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,2560,0.012672888735930124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,3584,0.014428445034556918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,3072,0.0124177775449223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,2048,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,3072,0.013145777914259167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,7168,0.030327998929553564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,2560,0.011077333655622272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,1536,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,1024,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,2560,0.012096000214417776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,8192,0.02758755617671543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,768,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,2048,0.011001778145631155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,1536,0.00853777759604984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,512,0.0042204442951414315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,1024,0.007680888805124495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,768,0.007245333658324347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,1536,0.009528888596428765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,256,0.003855111284388436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,768,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,512,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,128,0.003162666741344664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,256,0.0063999998900625445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,64,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,256,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,128,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,32,0.0031893334041039148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,128,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,64,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,32,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,65536,0.21611200438605416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,65536,0.1661333375506931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,51200,0.16807288593716094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,2048,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,16384,0.05821777714623345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,65536,0.12168799506293403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,51200,0.12785332732730442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,12288,0.04609333475430807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,51200,0.097397334045834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,16384,0.03529422150717841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,16384,0.04548888736300998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,10240,0.03825599948565165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,512,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,12288,0.028600000672870215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,12288,0.03485511077774896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,8192,0.03199377655982971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,7168,0.028061333629820082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,10240,0.0240631103515625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,10240,0.03053955568207635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,8192,0.020599111914634705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,6144,0.02424000038041009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,4096,0.015630222029156156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,7168,0.018927110566033255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,8192,0.02612888813018799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,5120,0.020648888415760465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,7168,0.02272622287273407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,4096,0.017873777283562552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,6144,0.018000889155599806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,3584,0.014390221900410123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,6144,0.02027555637889438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,5120,0.015306666493415833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,3072,0.012475555141766867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,5120,0.01758577757411533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,3584,0.013051555388503604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,2560,0.011032889286677042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,3584,0.014093332820468478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,3072,0.011924444801277585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,2048,0.009754666851626502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,3072,0.012960889273219638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,2560,0.010321777727868822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,1536,0.007576000359323289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,2560,0.011741333537631564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,1024,0.008295999632941352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,1024,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,1536,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,2048,0.010684444672531553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,768,0.004856889032655292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,1536,0.008946666287051307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,1024,0.007793777518802219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,1024,0.007559110720952352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,512,0.0038373333712418876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,768,0.0070293330483966405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,256,0.003140444349911478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,768,0.006990222467316522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,512,0.006895111252864202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,128,0.0028248888750871024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,512,0.006309333360857434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,256,0.006329777754015393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,4096,0.013869333598348828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,64,0.002512000087234709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,256,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,4096,0.01521511044767168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,128,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,128,0.005979555762476391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,65536,0.1791679991616143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,32,0.006561777657932705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,2048,0.008995555341243744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,64,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,51200,0.1396968894534641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,16384,0.04901866780387031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,65536,0.07168889045715332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,51200,0.05754399961895413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,12288,0.03912533322970072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,16384,0.035143110487196184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,16384,0.022985777921146814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,10240,0.0318853325313992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,12288,0.027088888817363318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,12288,0.018609777092933655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,8192,0.027457776996824477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,10240,0.016362667083740234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,32,0.0028719999310043124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,7168,0.023600889576805964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,6144,0.020455999506844413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,8192,0.014568888478808932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,7168,0.01346666696998808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,10240,0.023657777243190344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,5120,0.01757688820362091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,7168,0.017806222041447956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,6144,0.012527999778588613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,4096,0.014763555592960782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,8192,0.02010311186313629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,6144,0.016044444508022733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,5120,0.01146933353609509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,65536,0.1239466667175293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,3584,0.01314755529165268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,3072,0.011718221836619906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,4096,0.010417777631017896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,2560,0.010083555347389644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,4096,0.012535110943847232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,51200,0.09408266676796807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,5120,0.01426488823360867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,3072,0.008616000413894653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,2048,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,1536,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,2560,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,2048,0.007383999725182851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,2560,0.010031999813185798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,3072,0.011090666883521609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,1024,0.005737777799367905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,768,0.004567111117972268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,1536,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,1536,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,512,0.0038319999972979226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,2048,0.009144888983832465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,768,0.006511111226346757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,1024,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,1024,0.006664000037643645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,256,0.003129777808984121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,128,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,512,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,512,0.0063528890411059065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,256,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,256,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,128,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,128,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,32,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,3584,0.009347555538018545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,64,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,65536,0.16263200177086726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,32,0.006265777680608962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,3584,0.011692444483439127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,51200,0.13010133637322321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,65536,0.06878578000598483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,65536,0.12193244033389622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,16384,0.04661688870853848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,51200,0.09303022093243069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,12288,0.036664889918433294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,16384,0.02204177776972453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,16384,0.033999999364217125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,768,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,10240,0.03056266572740343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,8192,0.025588444537586633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,10240,0.015837333268589444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,64,0.0024986666523747975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,10240,0.022784888744354248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,7168,0.02274666726589203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,8192,0.014073777529928418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,6144,0.019595555133289762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,7168,0.017269333203633625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,7168,0.013166222307417126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,5120,0.016959110895792644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,6144,0.015831111205948722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,6144,0.012166221936543783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,4096,0.014628445108731588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,51200,0.05554844604598152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,4096,0.012482666307025485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,5120,0.0140328887436125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,12288,0.017948443690935772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,3584,0.011687999798191918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,12288,0.026111111044883728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,3584,0.008997333546479544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,3072,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,3072,0.010871999793582492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,2560,0.00941422250535753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,3584,0.011576000187132092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,2560,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,2048,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,2560,0.009742221898502773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,1536,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,2048,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,2048,0.008375999828179678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,5120,0.011039999624093374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,1024,0.005192888693677055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,1536,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,768,0.004512888689835866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,1536,0.00703022215101454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,4096,0.009686222506894005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,512,0.003926222109132343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,1024,0.00758133331934611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,1024,0.006679111056857639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,256,0.0031742221779293488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,768,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,128,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,512,0.006037333359320958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,256,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,64,0.0023004443695147834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,8192,0.01959111127588484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,32,0.002499555548032125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,128,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,128,0.005986666513813867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,256,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,64,0.005719111197524601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,65536,0.14317066139645046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,32,0.006309333360857434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,51200,0.11331733067830403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,16384,0.0424231125248803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,65536,0.06632266441980998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,51200,0.05388799972004361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,65536,0.11257155736287434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,12288,0.03483555714289347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,768,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,16384,0.03280444608794318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,3072,0.01036622209681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,16384,0.02128444446457757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,10240,0.029764443635940552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,12288,0.017273777061038546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,8192,0.02036711076895396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,7168,0.01776711146036784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,512,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,8192,0.013623110949993134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,10240,0.021652445197105408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,12288,0.024737776981459722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,8192,0.018851555056042142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,5120,0.013766222529941134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,7168,0.012749333348539142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,6144,0.015629332926538255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,6144,0.011912888950771756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,5120,0.010781333678298526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,4096,0.011301333705584208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,7168,0.016893333858913846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,3584,0.010328888893127441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,5120,0.013717333475748697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,4096,0.009148444566461775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,4096,0.012269333004951477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,51200,0.08850666549470688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,3584,0.008360889222886827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,3072,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,2560,0.00850399997499254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,3072,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,3072,0.010382222632567087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,2048,0.007430222299363878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,2560,0.009303111169073317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,2560,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,10240,0.015442666080262927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,1536,0.00655644428398874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,6144,0.015584000282817416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,2048,0.007153777612580194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,1024,0.004668444395065308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,1536,0.006811555474996567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,2048,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,768,0.004192000048028098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,1536,0.007102222078376346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,1024,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,1024,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,768,0.006325333482689328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,256,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,512,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,512,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,128,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,256,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,256,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,64,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,128,0.0057839999596277876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,32,0.002475555572244856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,64,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,32,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,65536,0.11410133043924968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,51200,0.0909315546353658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,512,0.0035386664999855887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,65536,0.10931733581754897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,65536,0.06371821959813435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,768,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,16384,0.03440622157520718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,12288,0.028175112273957994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,51200,0.051790220869912036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,51200,0.08474222156736587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,10240,0.024854222933451336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,16384,0.031054221921496924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,3584,0.011390222443474663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,8192,0.021528888079855178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,10240,0.015104000767072042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,128,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,7168,0.015607110328144498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,10240,0.020983111527231004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,8192,0.013247110777431063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,8192,0.018526222970750596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,7168,0.012368000215954252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,7168,0.01643022232585483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,5120,0.011996444728639392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,6144,0.011367999845080905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,4096,0.010629333555698395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,6144,0.014968888627158271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,5120,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,3584,0.01019200020366245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,4096,0.00906044410334693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,5120,0.013423999978436364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,3072,0.009630222287442949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,4096,0.011944889194435544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,3584,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,2560,0.008273777862389883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,3584,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,3072,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,12288,0.016869333055284288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,2048,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,3072,0.00981155534585317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,12288,0.024015110399987962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,1536,0.006162666612201267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,2560,0.00755555596616533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,1024,0.00453333349691497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,2560,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,1536,0.006807111203670502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,2048,0.007771555748250749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,768,0.003953777667548922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,1024,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,512,0.003540444291300244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,768,0.006227555374304454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,1024,0.006809777683681912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,256,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,768,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,512,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,128,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,512,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,256,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,64,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,256,0.005416888743638992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,32,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,128,0.005446222093370225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,16384,0.020524443851576913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,64,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,65536,0.09199021922217475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,32,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,51200,0.07451999849743314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,65536,0.06165422333611382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,2048,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,65536,0.06842666864395142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,16384,0.029205332199732464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,6144,0.013526221944226159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,51200,0.052998221582836576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,51200,0.049499554766549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,12288,0.02427733275625441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,16384,0.019564444820086162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,10240,0.021433777279324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,16384,0.020581333173645865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,12288,0.01640266676743825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,8192,0.015462223026487561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,10240,0.014440000057220459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,12288,0.016247999336984422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,7168,0.013577777478430005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,10240,0.01460622251033783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,8192,0.012967999610635968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,6144,0.01205244412024816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,128,0.005882666756709416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,8192,0.013167999684810638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,7168,0.011759999725553723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,5120,0.010414222048388587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,4096,0.00962311112218433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,7168,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,6144,0.01035733355416192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,3584,0.008175111479229396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,5120,0.00904088881280687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,6144,0.011142222417725457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,4096,0.008410666551854875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,5120,0.010168000227875179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,3072,0.007371555599901412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,4096,0.00924533357222875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,2560,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,3584,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,2048,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,3584,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,3072,0.007979555262459649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,1536,0.004596444467703502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,2560,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,1536,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,2560,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,1024,0.0035386664999855887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,2048,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,1536,0.006820444431569841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,768,0.0032097777972618737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,1536,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,1024,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,512,0.002867555452717675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,768,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,1024,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,256,0.0026977778308921387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,512,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,768,0.00563822231358952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,128,0.002456888970401552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,64,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,512,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,32,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,256,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,128,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,128,0.005072888814740711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,65536,0.08601778083377415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,64,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,32,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,51200,0.06900622447331746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,16384,0.027687110834651526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,65536,0.06043644746144613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,65536,0.0657822224828932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,12288,0.02201244400607215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,51200,0.04827022221353319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,3072,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,51200,0.05129422081841362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,16384,0.019472888774342008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,2048,0.007154666715198093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,10240,0.01958666741847992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,8192,0.015507555670208402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,16384,0.01958933307064904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,12288,0.016083555089102853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,7168,0.012851555314328937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,10240,0.01445777714252472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,6144,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,12288,0.01575288838810391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,8192,0.01275199982855055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,10240,0.014073777529928418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,8192,0.013000888956917657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,5120,0.010064889159467485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,7168,0.011367110742463006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,4096,0.009207111265924241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,6144,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,7168,0.011697777443461947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,3584,0.0079297779334916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,5120,0.009002666506502364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,256,0.005635555419656966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,3072,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,4096,0.008353778057628209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,5120,0.009767110976907942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,2560,0.00645955569214291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,4096,0.008386666576067606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,3584,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,2048,0.005479111025730769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,3072,0.007698666718271043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,3584,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,1536,0.004392888810899523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,2560,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,3072,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,2048,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,1024,0.003552000141806073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,2560,0.006908444480763541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,768,0.003160888950030009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,2048,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,1024,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,1536,0.006170666466156642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,512,0.0028000000036425064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,768,0.006211555666393704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,1024,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,256,0.0028106667515304354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,768,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,128,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,512,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,256,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,64,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,256,0.004984000076850255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,32,0.002209777840309673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,128,0.005694222119119432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,65536,0.07542666461732653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,6144,0.010774222513039907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,64,0.005162666655249066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,32,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,51200,0.0613280004925198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,65536,0.06318489048216078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,16384,0.023008000519540574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,51200,0.04731200138727824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,51200,0.04906222224235535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,1536,0.006773333168692059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,12288,0.019208000765906442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,16384,0.01917155583699544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,10240,0.016460445192125108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,12288,0.015506666567590503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,10240,0.013754666679435305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,8192,0.014396443963050842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,128,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,8192,0.012485333614879184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,8192,0.011763555308183035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,7168,0.013704000247849358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,65536,0.05951822466320462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,7168,0.010979555547237396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,7168,0.011330666641394297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,6144,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,6144,0.010605333579911126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,12288,0.01577333278126187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,512,0.005889777921968036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,4096,0.008613333106040955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,5120,0.009522666533788046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,10240,0.014141332772043018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,3584,0.007538666327794393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,4096,0.00835111074977451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,3072,0.006861333217885759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,3584,0.008067555725574493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,3584,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,2560,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,3072,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,3072,0.007264888948864407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,2048,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,2560,0.007280888656775157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,2560,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,1536,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,2048,0.0072266666425599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,2048,0.006464889066086874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,16384,0.019112888309690688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,1024,0.004237333519591225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,1536,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,768,0.0038497779104444715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,1536,0.00602311102880372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,5120,0.009045333498054082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,512,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,1024,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,768,0.006036444256703059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,256,0.0028328889360030494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,768,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,4096,0.00796977761718962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,512,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,512,0.005281777845488654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,128,0.002463999928699599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,64,0.002130666747689247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,256,0.005680888891220093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,32,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,256,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,128,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,128,0.00498933345079422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,6144,0.012476444244384766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,64,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,32,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,51200,0.052601777844958834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,65536,0.06085155407587687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,65536,0.058147556251949735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,16384,0.020631111330456205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,12288,0.01720177796151903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,51200,0.04600444436073303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,5120,0.009735110733244155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,51200,0.04730400111940172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,10240,0.014834667245546976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,16384,0.01871200071440803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,12288,0.015272888872358533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,1024,0.006562666760550604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,12288,0.015072888798183866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,8192,0.013051555388503604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,7168,0.012140444583363004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,10240,0.01316088851955202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,8192,0.011220444407727985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,10240,0.013435555828942193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,6144,0.011241777903503843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,8192,0.011960888902346293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,7168,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,5120,0.009633777870072259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,7168,0.01107911103301578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,6144,0.009705777797434065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,4096,0.008573333422342936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,6144,0.0099982221921285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,5120,0.008991999758614434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,3584,0.007532444265153673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,65536,0.0647511084874471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,5120,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,3072,0.006654222392373615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,4096,0.00829777783817715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,2560,0.0058577776783042485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,4096,0.007925333248244392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,3584,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,2048,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,3072,0.00774133370982276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,3584,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,1536,0.004230222354332606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,3072,0.007151111132568783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,2560,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,1024,0.003559111307064692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,2560,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,16384,0.018926221463415358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,1536,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,768,0.0032017777363459268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,2048,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,512,0.0028408887899584244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,1024,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,1536,0.006230222268237009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,1024,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,768,0.006036444256703059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,768,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,512,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,128,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,512,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,64,0.0021537778278191886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,256,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,128,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,32,0.002206222257680363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,128,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,65536,0.05317777726385328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,64,0.005037333402368757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,32,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,51200,0.04352533486154345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,65536,0.05682133303748237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,16384,0.01754755609565311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,51200,0.04515199859937032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,51200,0.04539555642339918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,2048,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,12288,0.0140337778462304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,16384,0.018547554810841877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,16384,0.01824355622132619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,10240,0.01239911135700014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,12288,0.014521777629852295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,256,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,8192,0.010752889017264048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,10240,0.012441777520709567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,7168,0.010423111418883005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,10240,0.013096888860066732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,8192,0.011007111105653973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,256,0.005629333357016246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,6144,0.009930666950013902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,5120,0.00905511114332411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,8192,0.011540444360838996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,7168,0.010325333310498131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,4096,0.007751999629868402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,7168,0.010629333555698395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,6144,0.009713778065310584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,3584,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,5120,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,6144,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,3072,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,5120,0.00871022211180793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,4096,0.008283555507659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,2560,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,3584,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,3072,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,65536,0.05780800183614095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,2048,0.004560889055331548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,2560,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,3072,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,1536,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,2560,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,1024,0.003216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,1536,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,2048,0.0063884444534778595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,1536,0.006085333310895496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,768,0.0029102222373088202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,1024,0.006324444380071428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,1024,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,12288,0.014725334114498563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,512,0.0025439999169773525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,768,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,256,0.0023431111541059283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,768,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,512,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,256,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,128,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,256,0.004960888789759742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,64,0.002130666747689247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,128,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,4096,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,32,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,128,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,65536,0.040920890039867826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,64,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,51200,0.03277333246337043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,32,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,2048,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,65536,0.055071999629338585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,16384,0.014550222290886773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,65536,0.05550133188565572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,12288,0.012553777959611682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,51200,0.04451555675930447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,16384,0.017775111728244357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,51200,0.04358222087224325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,10240,0.011394666300879585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,16384,0.01772444446881612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,12288,0.013445333474212222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,12288,0.014367999302016364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,512,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,7168,0.009728000395827824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,10240,0.012770666844315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,8192,0.011055111057228513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,8192,0.010654222634103564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,6144,0.009189333352777693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,7168,0.010135999984211391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,5120,0.008461332983440822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,7168,0.010009778042634329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,6144,0.00962933318482505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,4096,0.00776088900036282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,3584,0.007540444533030192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,6144,0.0092613332801395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,3584,0.0063991112013657885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,5120,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,5120,0.008593777815500895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,3072,0.005653333332803514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,4096,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,3584,0.008081778056091731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,2560,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,3584,0.007325333025720384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,3072,0.007679999702506595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,2048,0.004559999952713649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,3072,0.0070186663005087115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,2560,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,2560,0.006708444406588872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,8192,0.010503999888896942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,2048,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,1024,0.0034391110142072043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,10240,0.012431999875439538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,2048,0.006528888725572162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,768,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,1536,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,512,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,1024,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,1536,0.006015111174848344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,256,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,1024,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,512,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,768,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,128,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,256,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,512,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,64,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,32,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,256,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,128,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,65536,0.03019022279315525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,128,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,51200,0.02439022229777442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,64,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,4096,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,32,0.005004444470008214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,16384,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,65536,0.05461333195368449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,65536,0.052777777115503945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,12288,0.012717333104875354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,51200,0.044300443596310086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,16384,0.016174222032229107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,51200,0.042062222957611084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,12288,0.013416888813177744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,1536,0.004159111115667555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,8192,0.010301333334710863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,12288,0.013422222601042854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,10240,0.012072888513406118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,7168,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,10240,0.011752000285519494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,8192,0.010764444039927589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,6144,0.00980000032318963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,8192,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,768,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,7168,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,7168,0.009922666682137383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,6144,0.009659555223253038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,4096,0.0068915556702348925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,6144,0.009204444785912832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,5120,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,3584,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,5120,0.008637333081828224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,4096,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,3072,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,4096,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,3584,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,2560,0.005020444591840108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,3584,0.007365333537260692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,3072,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,2048,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,10240,0.011369778050316704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,2560,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,16384,0.01721244388156467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,2560,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,2048,0.0068986668354935115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,1024,0.0037848887344201407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,2048,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,1536,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,768,0.0034426665968365143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,1536,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,1024,0.006292444550328785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,5120,0.008354666332403818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,512,0.0028159999185138275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,1024,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,768,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,256,0.00250133333934678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,768,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,512,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,256,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,512,0.0053013331360287136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,64,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,256,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,32,0.0022008888837363985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,128,0.005543111099137201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,128,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,65536,0.023383999864260357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,64,0.004953777624501122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,32,0.004998222407367495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,51200,0.02049688829316033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,3072,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,65536,0.05475733346409268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,16384,0.012562666502263812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,1536,0.0042088888585567474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,65536,0.05165333218044705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,12288,0.010692444112565784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,51200,0.04366133279270596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,10240,0.010115555591053432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,51200,0.04166933231883579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,12288,0.013455111119482251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,16384,0.016137777103318107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,8192,0.008802666432327693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,12288,0.013051555388503604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,7168,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,10240,0.011759999725553723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,8192,0.010696888797812991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,6144,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,8192,0.010346666806273991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,128,0.0021084443562560612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,5120,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,7168,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,6144,0.009391110804345872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,4096,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,5120,0.008848889006508721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,6144,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,3584,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,5120,0.0085235552655326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,4096,0.00816888858874639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,3072,0.0057546666098965546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,3584,0.007715555528799693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,2560,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,3584,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,3072,0.0075662218862109715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,2048,0.004513777792453766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,3072,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,2560,0.007354666789372762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,1536,0.0038808890514903595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,2560,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,16384,0.015867556134859722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,2048,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,1024,0.0031617778456873367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,2048,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,1536,0.006545777950021956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,768,0.002855111120475663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,10240,0.012072888513406118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,1536,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,512,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,768,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,1024,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,256,0.0022728888110982045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,768,0.005504000104135937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,7168,0.010020444790522257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,128,0.0021484444538752237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,512,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,256,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,256,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,64,0.0018959999498393799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,768,32,0.001863111017478837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,128,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,4096,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,64,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,32,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,65536,0.0532897777027554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,65536,0.051013333929909595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,16384,0.010109333528412713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,51200,0.04038844505945841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,51200,0.04111822115050422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,12288,0.008632889224423302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,16384,0.015834665960735746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,1024,0.006090666684839461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,16384,0.01587377819750044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,10240,0.007825777762466008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,12288,0.013239111337396832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,512,0.005632888939645555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,8192,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,10240,0.011925333076053195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,7168,0.00853777759604984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,10240,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,8192,0.01074488874938753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,6144,0.009191110730171204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,8192,0.010374222364690568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,7168,0.010276444256305695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,65536,0.018888889087571036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,5120,0.00796977761718962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,7168,0.009686222506894005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,4096,0.006814222368929122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,51200,0.015082667271296183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,5120,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,3584,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,5120,0.008319999608728621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,4096,0.00832088871134652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,4096,0.007615110940403408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,3072,0.005535111245181825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,3584,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,2560,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,3584,0.007370666497283512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,3072,0.007676444119877285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,2048,0.004238222208287981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,3072,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,2560,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,1536,0.0038364442686239877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,2560,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,12288,0.0129457778400845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,2048,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,1024,0.0031662223239739737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,2048,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,768,0.0028408887899584244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,1536,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,512,0.002516444358560774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,1024,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,1536,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,256,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,768,0.0058853332367208265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,1024,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,128,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,6144,0.009377777576446533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,256,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,512,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,6144,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,64,0.0019662222928471034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,512,32,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,65536,0.014966222974989148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,128,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,64,0.005166222237878376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,32,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,51200,0.013340444200568728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,65536,0.04686577783690559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,65536,0.04973511232270134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,16384,0.008993777963850234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,51200,0.03791200121243795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,12288,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,51200,0.039148443275027804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,16384,0.015832000308566622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,10240,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,12288,0.01311555587583118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,12288,0.01310400002532535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,8192,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,10240,0.011752888560295105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,512,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,7168,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,8192,0.010715555813577441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,8192,0.01036888857682546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,256,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,6144,0.007558222446176741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,128,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,7168,0.010079111489984725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,5120,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,7168,0.00965244405799442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,6144,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,4096,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,5120,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,6144,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,3584,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,5120,0.008291555775536431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,4096,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,3072,0.0053048887186580235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,4096,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,16384,0.015803555647532146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,2560,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,3584,0.007300444775157505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,3072,0.007515555454625024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,2048,0.004217777815130022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,3072,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,2560,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,1536,0.003552888830502828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,2560,0.006599111275540457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,2048,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,1024,0.0030862221287356485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,768,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,2048,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,1536,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,768,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,1024,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,512,0.002467555511328909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,768,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,1024,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,256,0.0021928888228204516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,768,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,512,0.005731555736727185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,128,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,512,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,256,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,64,0.0018542222678661346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,256,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,256,32,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,128,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,3584,0.007714666426181793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,128,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,65536,0.013117333253224691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,51200,0.011590222517649332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,64,0.004821333206362194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,32,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,65536,0.04652533266279432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,16384,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,65536,0.04719110992219713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,12288,0.006993778049945831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,51200,0.03802311089303758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,10240,0.011747555600272285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,10240,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,16384,0.015537778536478678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,12288,0.013367999758985309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,1536,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,8192,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,12288,0.012815111213260226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,10240,0.012055111428101858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,7168,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,10240,0.011399110986126794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,6144,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,7168,0.010031999813185798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,7168,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,5120,0.0064693333374129395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,6144,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,6144,0.008951110972298516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,4096,0.005647999958859549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,5120,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,3584,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,4096,0.008173333273993598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,4096,0.007449777589903937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,3072,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,51200,0.03765689002143012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,3584,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,2560,0.004872888740566042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,16384,0.015840000576443143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,3584,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,2048,0.004226666771703296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,3072,0.007367110914654202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,3072,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,1536,0.003499555504984326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,2048,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,2560,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,1024,0.002867555452717675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,2048,0.006070222291681502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,8192,0.0107342220014996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,1536,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,768,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,1536,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,8192,0.01033688916100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,1024,0.006070222291681502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,512,0.0024702221983008915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,256,0.002114666625857353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,1024,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,768,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,128,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,512,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,768,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,256,0.00563733321097162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,512,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,32,0.0018035556293196147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,128,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,256,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,65536,0.011687111523416309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,128,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,51200,0.010104888843165504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,64,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,16384,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,32,0.005048888838953442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,12288,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,65536,0.04712177647484673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,10240,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,51200,0.03780444463094076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,2560,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,8192,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,7168,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,12288,0.012801777985360889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,10240,0.011437333292431302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,6144,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,8192,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,5120,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,4096,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,7168,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,3584,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,5120,0.008368000388145447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,3072,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,128,64,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,4096,0.0076248885856734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,2560,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,2048,0.00453244439429707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,5120,0.008390222158696916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,1536,0.0034977777136696708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,3584,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,1024,0.0028702221396896574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,3072,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,768,0.0025066667132907445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,2560,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,512,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,1536,0.005683555371231503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,256,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,1024,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,128,0.0018257778137922287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,16384,0.015855110353893705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,512,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,64,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,64,32,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,256,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,65536,0.012083555261294046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,128,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,51200,0.01016088906261656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,64,0.004949333353175057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,16384,0.008352888955010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,32,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,12288,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,6144,0.008978666530715095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,65536,0.04699022240108914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,51200,0.03780444463094076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,8192,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,16384,0.015800888339678448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,7168,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,12288,0.013040000365840064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,6144,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,5120,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,10240,0.011394666300879585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,4096,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,2048,0.005975111077229182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,8192,0.010327999790509542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,3584,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,7168,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,3072,0.004950222041871813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,6144,0.009011555876996782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,2560,0.004867555366622077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,2048,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,5120,0.008351999852392409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,768,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,1536,0.003510222252872255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,4096,0.007458666960398356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,1024,0.002871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,768,0.0025928889711697898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,3584,0.007222221957312689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,3072,0.006681777950790193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,512,0.0024560000747442245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,2560,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,256,0.0021199999998013177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,128,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,2048,0.005972444597217772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,1536,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,64,0.0018071112119489247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,1024,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,32,0.0017884444031450485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,768,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,512,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,256,0.004959110998445087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,128,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,97,32,10240,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,64,0.004951111144489712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,32,0.004596444467703502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,16384,0.37219821082221133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,16384,0.2025324503580729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,16384,0.13839911090003118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,51200,0.605055120256212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,12288,0.2782844437493218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,12288,0.14949688646528456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,51200,1.1494053734673393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,12288,0.10817688703536987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,10240,0.20877333482106528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,51200,0.400851567586263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,10240,0.12643644544813368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,10240,0.08477333519193862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,8192,0.1732808881335788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,8192,0.07303822040557861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,8192,0.1072453326649136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,12288,0.2205982208251953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,7168,0.15049778090582952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,7168,0.09047289027108087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,10240,0.18802666664123535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,6144,0.12872621748182508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,7168,0.062216891182793506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,8192,0.14998932679494223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,6144,0.07910666863123576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,6144,0.05712888638178507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,5120,0.1091466678513421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,7168,0.13204533523983425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,5120,0.06624355581071642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,4096,0.09173333644866943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,5120,0.045590221881866455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,6144,0.11317243840959336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,4096,0.038711110750834145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,4096,0.05453244513935513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,3584,0.07880889044867621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,5120,0.09636711412005955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,3072,0.0685173339313931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,4096,0.07949333058463202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,3584,0.03717155588997735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,2560,0.05743555227915446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,3072,0.04270311196645101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,3072,0.031423999203575984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,3072,0.06427288717693753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,3584,0.07241421937942505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,2560,0.037019554111692644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,2048,0.04723111126157972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,16384,0.28668178452385795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,1536,0.036714666419559054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,2048,0.023172444767422144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,2048,0.03213511241806878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,1536,0.019143111175960965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,1536,0.025320000118679468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,2048,0.04894933435651991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,1024,0.027106665902667578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,1536,0.041138665543662176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,768,0.020279111133681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,1024,0.01551822159025404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,1024,0.025036444266637165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,768,0.013815999858909182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,1024,0.03390844331847297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,512,0.01407199932469262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,768,0.020738666256268818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,2560,0.02776177724202474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,256,0.00832088871134652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,768,0.02958400050799052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,512,0.015472888946533203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,51200,0.9205759896172417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,2560,0.05684355232450697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,256,0.012327999704413943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,128,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,256,0.011297778122954898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,64,0.004811555561092165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,128,0.015602666470739575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,128,0.011360000405046674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,32,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,3584,0.048899554544024997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,128,0.016146666473812528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,64,0.011420444481902652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,32,0.015454222758611044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,512,0.012358222570684222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,512,0.025010666913456384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,65536,1.0958710776435006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,256,0.020617778102556866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,51200,0.4797813097635905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,16384,0.2774346669514974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,65536,0.4301022158728705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,16384,0.16162755754258898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,12288,0.20868444442749023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,65536,0.6276319821675619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,51200,0.8603119850158691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,12288,0.11877155303955078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,51200,0.32654666900634766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,10240,0.17177066538068983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,12288,0.0854826635784573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,10240,0.10403644376330906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,16384,0.11174578136867946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,12288,0.17828533384535047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,10240,0.06823022498024835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,8192,0.138483550813463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,7168,0.1208311054441664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,8192,0.08572977781295776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,8192,0.05761510796017117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,10240,0.14889955520629883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,6144,0.10478133625454372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,7168,0.07253155443403456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,7168,0.05030133326848348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,8192,0.12049955791897243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,6144,0.06337777773539226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,6144,0.04664088951216804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,16384,0.24003643459743926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,5120,0.08888800276650323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,7168,0.10711733500162761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,5120,0.05338222119543287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,6144,0.09494755665461223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,5120,0.03829688827196757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,4096,0.07302577628029717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,3584,0.06298844681845771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,5120,0.08118933439254761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,4096,0.03278133273124695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,4096,0.04440088735686409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,3072,0.05503822366396586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,4096,0.06936177942487928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,3584,0.03990222348107232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,3072,0.03490044342146979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,51200,0.716669347551134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,3584,0.029375112719006006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,2560,0.04644088943799337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,3072,0.025827555192841425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,2048,0.03781422310405307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,2560,0.023344000180562336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,2560,0.035232888327704534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,65536,0.9346746868557401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,2560,0.04885777831077576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,1536,0.030928889910380047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,2048,0.020615110794703167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,1536,0.017072000437312655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,2048,0.04302755660480923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,1536,0.021055110626750525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,1024,0.021772444248199463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,1024,0.01370933320787218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,768,0.016037333342764113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,1024,0.020929776959949072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,3584,0.06199466519885593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,1536,0.036914666493733726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,512,0.01128355579243766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,1024,0.029836446046829224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,768,0.0123831108212471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,768,0.017747556169827778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,256,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,512,0.013418667018413544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,3072,0.056194669670528836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,2048,0.026783999469545152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,768,0.02570488883389367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,128,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,512,0.022101332743962605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,256,0.009885333478450775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,64,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,128,0.010024000373151567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,32,0.004257777912749184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,128,0.00997688869635264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,256,0.018586667047606576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,64,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,128,0.014298665854665967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,32,0.013209777573744455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,65536,0.3415493435329861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,65536,0.22648888164096406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,51200,0.26757778061760795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,65536,0.14672356181674534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,51200,0.17354133394029406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,16384,0.0897573365105523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,512,0.01096088853147295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,12288,0.06936444176567926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,51200,0.11860444810655381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,256,0.01075111081202825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,16384,0.04319999946488274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,16384,0.06106133593453301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,10240,0.056688891516791455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,12288,0.03503377901183234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,12288,0.04601600103908115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,16384,0.06111022498872545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,8192,0.0473306675752004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,10240,0.039649777942233615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,10240,0.028543111350801256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,12288,0.047192000680499606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,7168,0.0414631101820204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,8192,0.0343777769141727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,8192,0.02535999980237749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,6144,0.03607822126812405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,7168,0.02982577681541443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,7168,0.021517333057191636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,65536,0.2045768896738688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,51200,0.15918933020697698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,5120,0.030765334765116375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,7168,0.032658666372299194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,6144,0.020288000504175823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,6144,0.02637600070900387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,4096,0.025217778152889673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,5120,0.018238221605618794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,5120,0.023023999399609033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,3584,0.022662222385406494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,6144,0.029951999584833782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,4096,0.01611288885275523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,4096,0.019384000036451552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,5120,0.02597066594494714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,3072,0.019890666007995605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,3584,0.014192889134089151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,10240,0.04126844472355313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,3072,0.013083555632167392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,3072,0.015538665983412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,3584,0.021585777401924133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,2560,0.0166302224000295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,3072,0.019380443625979953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,8192,0.03616355525122748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,2560,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,2560,0.013903111219406128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,2560,0.018087999688254464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,2048,0.011493333511882357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,2048,0.012444444828563266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,1536,0.010877333581447601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,2048,0.0166275550921758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,1536,0.009647111097971598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,1536,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,1536,0.01553066737122006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,4096,0.023578666978412207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,1024,0.010667555862002902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,768,0.007237333390447829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,1024,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,1024,0.013792888985739814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,512,0.0058204444746176405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,768,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,768,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,2048,0.014121777481502958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,256,0.003572444534964032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,512,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,768,0.012429333395428128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,512,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,128,0.0031724443866146933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,256,0.006654222392373615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,256,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,512,0.011011555790901184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,64,0.0028071111689011254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,1024,0.008808888494968414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,128,0.006867555694447623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,128,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,32,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,256,0.01035644445154402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,64,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,128,0.008858666651778752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,3584,0.017490666773584154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,32,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,65536,0.27170666058858234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,65536,0.17413422796461317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,51200,0.21069155799018013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,16384,0.07008177704281278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,51200,0.13325956132676867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,65536,0.12763999568091497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,12288,0.05470222234725952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,16384,0.0476435555352105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,16384,0.03741866681310866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,10240,0.04494577646255493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,12288,0.03648533423741659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,12288,0.029719998439153034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,16384,0.05184355709287855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,8192,0.03831111060248481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,12288,0.04132089018821716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,10240,0.025524444050259058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,10240,0.031649778286616005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,7168,0.033258666594823204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,51200,0.1322071154912313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,65536,0.17152622010972765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,8192,0.021884444687101576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,10240,0.036554667684766985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,6144,0.02872177627351549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,8192,0.027173333697848853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,7168,0.019873778025309246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,7168,0.02401777770784166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,5120,0.024756444825066462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,6144,0.018938667244381376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,6144,0.02148088812828064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,8192,0.032574223147498235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,4096,0.021293333835071985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,7168,0.029277334610621136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,51200,0.10116533438364665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,5120,0.016851555969980028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,3584,0.018238221605618794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,6144,0.026706667410002813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,5120,0.018079110317760043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,3072,0.01593866613176134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,4096,0.014535110857751636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,4096,0.01568000018596649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,3584,0.013306666579511432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,5120,0.023842665884229872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,3584,0.014504889647165934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,3072,0.012654222548007965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,3072,0.013094222380055321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,2048,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,3584,0.01959822244114346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,2560,0.011633777783976661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,2560,0.01204622205760744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,1536,0.009473777479595607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,2048,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,2560,0.01706133286158244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,2048,0.01088711122671763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,1024,0.0075911109646161394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,1536,0.008640888664457533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,2048,0.01572177807490031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,1024,0.007939555578761632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,1024,0.008555555509196388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,1536,0.014557333456145393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,768,0.006364444477690591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,1024,0.013152889079517789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,768,0.007664889097213745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,2560,0.013341333303186627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,512,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,768,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,4096,0.021120000216695998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,768,0.011396444506115384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,256,0.003481777591837777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,512,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,512,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,128,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,256,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,256,0.006048888795905643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,512,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,64,0.00254488881263468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,256,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,128,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,128,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,32,0.002628444383541743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,64,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,128,0.008005333443482717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,32,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,65536,0.2426462173461914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,65536,0.16536000039842394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,51200,0.19244799349043104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,65536,0.12065688769022624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,51200,0.12756889396243626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,16384,0.05894577503204346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,3072,0.018241778016090393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,51200,0.09864622354507446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,16384,0.04564355479346382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,12288,0.046241776810752026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,16384,0.03652088840802511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,1536,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,16384,0.04986222253905403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,10240,0.037984000311957464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,12288,0.02916711237695482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,12288,0.034879999028311834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,65536,0.15709600183698866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,8192,0.03219733304447598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,10240,0.02432800001568264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,12288,0.03991555505328708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,7168,0.02854222059249878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,8192,0.026018665896521673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,8192,0.02166044380929735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,10240,0.035599999957614474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,6144,0.02480711042881012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,8192,0.030897776285807293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,7168,0.019345778557989333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,5120,0.021634666456116572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,6144,0.020104888412687514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,6144,0.01759111053413815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,7168,0.027507555153634813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,4096,0.018238221605618794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,6144,0.02586488922437032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,5120,0.01741422216097514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,4096,0.014362666341993542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,4096,0.01517777807182736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,3584,0.013960889644092984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,5120,0.022853333089086745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,51200,0.12526222070058188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,3072,0.012583110895421771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,4096,0.020284444093704224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,10240,0.030352890491485596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,3584,0.013106666505336761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,2560,0.010995555255148145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,3584,0.013993778162532382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,3072,0.012237333589129977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,3072,0.013000888956917657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,2048,0.009398221969604492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,3584,0.01904266741540697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,2560,0.01073688848151101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,2560,0.011784888803958893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,1536,0.007920000288221572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,3072,0.017841777867741056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,2048,0.009214222431182861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,2560,0.016466667254765827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,1024,0.006137777947717243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,2048,0.010677333507272931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,1536,0.00886577781703737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,5120,0.015600888265503777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,2048,0.015613334046469795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,768,0.0047004446387290955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,1536,0.014108444253603617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,1024,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,1024,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,512,0.003843555433882607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,768,0.007104888558387756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,768,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,1024,0.011759111450778114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,256,0.0032471112079090546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,512,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,768,0.011734222372372946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,128,0.0028604444944196274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,512,0.010689777632554373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,256,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,256,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,7168,0.022675555613305833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,256,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,128,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,64,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,32,0.002481777841846148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,65536,0.18180621994866264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,64,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,1536,0.008290666672918532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,65536,0.12285778257581924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,65536,0.07132889164818658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,51200,0.13816000355614558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,51200,0.09460710816913181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,128,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,16384,0.05101066827774048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,65536,0.14071555932362875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,51200,0.058182219664255776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,128,0.008301333420806462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,32,0.0064151109092765385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,12288,0.03748088743951585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,16384,0.022944000032213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,16384,0.0349333319399092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,10240,0.030995554394192163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,51200,0.11216799418131511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,12288,0.01861422260602315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,12288,0.02718933257791731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,16384,0.04557955596182081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,8192,0.026421333352724712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,10240,0.023471999499532912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,10240,0.01629155543115404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,12288,0.03721333212322659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,7168,0.022886221607526142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,8192,0.02032444377740224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,8192,0.014575999644067554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,10240,0.03258311086230808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,6144,0.019895111521085102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,7168,0.013445333474212222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,7168,0.017627555463049147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,8192,0.029119110769695703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,5120,0.017031111650996737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,6144,0.012545777691735161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,6144,0.015989333391189575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,7168,0.02608888844648997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,4096,0.014662222729788886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,5120,0.014074666632546319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,6144,0.024528889192475214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,5120,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,3584,0.012078222301271228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,4096,0.01258933295806249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,4096,0.010690666735172272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,5120,0.021712889273961384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,3072,0.011029333704047732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,3584,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,2560,0.009690666364298927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,3584,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,4096,0.01956533392270406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,3072,0.008656000097592672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,2048,0.008256889051861232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,3072,0.010928000013033548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,3584,0.0184551113181644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,512,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,2560,0.008234666453467475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,1536,0.006647999915811751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,2560,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,3072,0.017842666970358956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,2048,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,1024,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,2048,0.008971555365456475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,2560,0.016511999898486666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,1536,0.007043555378913879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,768,0.004578666554556953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,1536,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,1024,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,2048,0.015093333191341825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,512,0.0032471112079090546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,1024,0.00721866637468338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,768,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,1536,0.013075555364290873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,256,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,768,0.006801777829726537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,1024,0.011703111231327057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,512,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,128,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,768,0.011381333072980246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,256,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,64,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,256,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,32,0.002327999927931362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,512,0.010048000348938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,128,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,256,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,128,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,64,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,65536,0.1645475493537055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,128,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,32,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,65536,0.11960800488789876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,51200,0.12957599427964953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,65536,0.06861422459284465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,16384,0.04791555470890469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,51200,0.05556533071729872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,16384,0.021682666407691106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,16384,0.03398311138153076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,12288,0.03471466567781236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,512,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,16384,0.04495733314090305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,51200,0.11220533317989773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,10240,0.028784000211291846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,12288,0.026086222794320848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,12288,0.017955554856194388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,8192,0.024659555819299486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,10240,0.015836444165971544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,10240,0.022452443838119507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,8192,0.014098667436175875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,8192,0.01959466603067186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,7168,0.021432888176706102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,8192,0.028516444895002577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,51200,0.09256177478366429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,7168,0.017277333471510146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,6144,0.018571555614471436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,7168,0.013082666529549493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,7168,0.02565777798493703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,5120,0.015806222955385845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,6144,0.015835555063353646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,65536,0.13924533791012234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,6144,0.023710222707854375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,5120,0.011008000208271874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,4096,0.013455111119482251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,5120,0.014142221874660917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,12288,0.036836445331573486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,3584,0.011608888705571493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,4096,0.009756444229020013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,5120,0.021770666042963665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,4096,0.01273333364062839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,10240,0.032587554719712995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,3072,0.010397333237859938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,3584,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,3584,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,4096,0.019344889455371432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,2560,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,3072,0.008303110798199972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,3072,0.010689777632554373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,2048,0.007881777981917063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,3584,0.0184542222155465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,2560,0.009858667022652095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,2560,0.00793333351612091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,3072,0.017088888419999015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,1536,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,6144,0.012025777664449481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,2048,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,2560,0.01587466730011834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,2048,0.008277333445019191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,1024,0.00500177757607566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,1536,0.0075493330756823225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,2048,0.015085332923465304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,1024,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,1536,0.013608000344700284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,1024,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,512,0.003547555456558863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,768,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,1024,0.011408888631396823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,256,0.0028026666906144884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,512,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,512,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,768,0.010416888528399997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,128,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,256,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,512,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,64,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,128,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,256,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,32,0.002247111043996281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,64,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,65536,0.13083467218610975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,128,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,32,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,1536,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,768,0.0038373333712418876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,51200,0.10364977518717448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,65536,0.11440000269148086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,65536,0.06605599986182319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,16384,0.03997777899106344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,768,0.006721777634488211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,51200,0.08804888857735528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,51200,0.05385422375467089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,12288,0.031962666246626116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,16384,0.032335999939176775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,16384,0.020971554848882887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,10240,0.027309333284695942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,65536,0.13892800278133816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,51200,0.11106133460998535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,12288,0.01737155516942342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,12288,0.02520888878239526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,16384,0.04462755719820658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,128,0.00534400012758043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,10240,0.015293333265516492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,10240,0.02165511084927453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,12288,0.03637955586115519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,8192,0.013592888911565145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,8192,0.019277332557572257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,10240,0.031832890378104314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,6144,0.017688888642523024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,7168,0.016957332690556843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,8192,0.027680888772010803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,6144,0.011704888608720569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,7168,0.025274667474958632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,5120,0.014505777094099255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,6144,0.015323556131786771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,6144,0.023353776997990076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,5120,0.01071999967098236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,5120,0.013776889277829064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,8192,0.022673777408070032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,4096,0.012141333685980903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,3584,0.011447110937701331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,4096,0.009354666703277165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,7168,0.019111999207072787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,4096,0.019176888797018263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,3072,0.009930666950013902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,3584,0.008642666869693333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,3584,0.011286222272449069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,7168,0.012738666600651212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,2560,0.008897778060701158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,3072,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,3584,0.018198221921920776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,2560,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,2560,0.009315555294354757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,3072,0.016703110602166917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,2048,0.00793511089351442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,4096,0.012083555261294046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,2048,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,1536,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,2560,0.015673778123325773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,2048,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,1024,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,5120,0.021399999658266704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,1536,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,2048,0.013906665974193148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,768,0.0042035554846127825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,1024,0.006736889067623351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,1024,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,1536,0.012737777498033313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,512,0.0031884445084465873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,1024,0.011381333072980246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,768,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,3072,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,512,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,512,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,256,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,768,0.010382222632567087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,128,0.002465777720014254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,256,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,512,0.010427555276287926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,256,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,64,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,32,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,128,0.005974222388532426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,128,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,256,0.009339555270142024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,65536,0.11382488409678142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,64,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,32,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,128,0.007668444679843054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,1536,0.007080000307824876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,51200,0.09086844656202529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,65536,0.10867110888163249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,16384,0.03460977805985345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,65536,0.06405511167314318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,51200,0.05147733290990194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,768,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,51200,0.08462399906582302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,12288,0.02913244565327962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,16384,0.03138755427466498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,65536,0.13825599352518717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,10240,0.02535111043188307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,12288,0.016689777374267578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,51200,0.11070044173134698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,12288,0.023959111836221483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,16384,0.044292443328433566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,8192,0.01832711034350925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,10240,0.015037332971890768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,12288,0.03578311204910278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,10240,0.02102488941616482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,8192,0.013266666895813413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,8192,0.018616000811258953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,6144,0.013639111485746173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,7168,0.01237777786122428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,8192,0.027278222971492346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,7168,0.016520000166363187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,5120,0.012000888586044312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,6144,0.015050666199790107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,6144,0.011067555182509951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,7168,0.0248346659872267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,4096,0.010681777364677854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,5120,0.010000000397364298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,5120,0.013398222625255585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,16384,0.0203146669599745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,4096,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,3584,0.010755555497275459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,4096,0.011874666644467248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,5120,0.02067733307679494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,3584,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,4096,0.019367110398080613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,3584,0.011100444528791638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,7168,0.015615999698638916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,3584,0.017853332890404593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,2560,0.008681777450773451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,3072,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,10240,0.0315048893292745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,3072,0.010327999790509542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,2048,0.007195555501514011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,2560,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,3072,0.017180444465743173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,2560,0.008369777765538957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,1536,0.006495111104514863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,2048,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,2048,0.007080888582600488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,2560,0.015146666102939181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,6144,0.02288533416059282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,1024,0.004831999954250124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,1536,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,1536,0.0069982219073507525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,768,0.004233777936961916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,1024,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,1024,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,1536,0.012804444465372296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,512,0.0031724443866146933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,3072,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,768,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,768,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,256,0.002871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,512,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,512,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,768,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,128,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,256,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,256,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,512,0.009674666656388177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,64,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,128,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,256,0.009008888569143083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,128,0.005430222385459476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,32,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,65536,0.09043200148476495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,128,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,64,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,32,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,2048,0.013452444639470843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,65536,0.06715555323494805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,51200,0.07238755623499553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,16384,0.028800888193978205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,1024,0.011051555474599203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,51200,0.05276088913281759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,12288,0.023427555958429973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,65536,0.13818932904137507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,16384,0.020464888877338834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,16384,0.019741333193249173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,10240,0.020640000700950623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,51200,0.11043466462029351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,12288,0.016185777054892648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,12288,0.016328000360065036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,16384,0.04365333252482944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,10240,0.014721777704026965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,10240,0.014710222681363424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,12288,0.035001777940326266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,7168,0.013321778012646569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,8192,0.013405332962671915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,8192,0.012706666356987424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,7168,0.011538666983445486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,7168,0.012073777616024017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,6144,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,8192,0.02660977840423584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,51200,0.04941244588957893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,7168,0.0247706671555837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,6144,0.011048888994587792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,5120,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,6144,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,4096,0.00943377779589759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,5120,0.009153777526484596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,6144,0.022336888644430373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,5120,0.010071111222108206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,3584,0.00814577771557702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,4096,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,4096,0.00903200027015474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,5120,0.0203493336836497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,3072,0.007416889071464539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,3584,0.008078222473462423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,3584,0.008271999657154083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,4096,0.0188711103465822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,10240,0.03080444534619649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,3072,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,3584,0.01732711162832048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,65536,0.06188888682259453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,2560,0.0069982219073507525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,3072,0.015842666228612263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,2048,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,2048,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,1536,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,2048,0.0064382221963670515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,2048,0.013463111387358772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,8192,0.015483554866578845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,1536,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,1024,0.003563555578390757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,1536,0.012051555845472546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,2560,0.006518222391605377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,3072,0.007755555212497711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,1024,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,768,0.0032559999575217566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,1024,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,2560,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,1024,0.010703999963071613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,768,0.006070222291681502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,768,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,2560,0.014259555273585849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,768,0.01034577770365609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,512,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,512,0.00553688903649648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,256,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,128,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,1536,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,256,0.009671111073758867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,128,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,64,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,128,0.007408888803588019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,32,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,64,0.0052266667286554975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,512,0.0028408887899584244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,65536,0.08166399929258558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,32,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,51200,0.06610044505861071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,65536,0.06532355811860827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,65536,0.060467554463280566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,512,0.0103493332862854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,256,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,16384,0.02693333393997616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,51200,0.051129778226216636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,51200,0.04871822065777249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,65536,0.13805333773295084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,12288,0.02163377735349867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,16384,0.019467555814319186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,128,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,16384,0.019837333096398246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,51200,0.11019466982947455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,12288,0.015802666544914246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,10240,0.01772533357143402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,12288,0.016181333197487723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,8192,0.017208000024159748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,16384,0.0437448885705736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,10240,0.014357333381970724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,12288,0.03432888785998026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,7168,0.012772444221708508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,10240,0.014195554786258273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,8192,0.01238933371173011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,8192,0.01290933373901579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,6144,0.01130222198035982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,10240,0.03052355514632331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,7168,0.011363555159833698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,7168,0.011633777783976661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,8192,0.02690222197108799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,5120,0.009933333430025313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,6144,0.010839111275143094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,4096,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,5120,0.009674666656388177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,5120,0.009853333234786987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,6144,0.022844443718592327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,3584,0.00776533368561003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,5120,0.02088622252146403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,3584,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,4096,0.018244443668259513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,3072,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,3584,0.007948444121413762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,3584,0.017252445220947266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,3072,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,2560,0.006335999816656113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,3072,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,256,0.002463999928699599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,3072,0.015805333852767944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,6144,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,2048,0.0053280000057485365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,7168,0.024069334069887798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,2560,0.006788444601827198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,1536,0.004353777815898259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,2048,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,2560,0.014782221780882942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,2048,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,1024,0.0035377778112888336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,1536,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,1536,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,2048,0.013575110998418597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,768,0.0031902222997612427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,1024,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,1536,0.011752888560295105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,768,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,768,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,512,0.00282222218811512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,1024,0.011396444506115384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,512,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,256,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,512,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,2560,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,512,0.01034755590889189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,256,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,128,0.0022053333620230355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,4096,0.008354666332403818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,256,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,128,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,128,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,64,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,1024,0.0058764442801475525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,32,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,65536,0.0690480007065667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,128,0.00738666703303655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,64,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,32,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,51200,0.05695911248524984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,768,0.01034400032626258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,65536,0.06291111310323079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,65536,0.05946133534113566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,16384,0.02092355489730835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,51200,0.04724355538686117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,65536,0.1380266613430447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,51200,0.04912622107399834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,256,0.005850666513045629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,12288,0.01719555589887831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,4096,0.008619555168681675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,16384,0.019308444526460435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,16384,0.019293333093325298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,10240,0.014727110664049784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,51200,0.10989866654078166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,12288,0.015838222371207345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,12288,0.015138667490747241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,8192,0.013665777941544851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,16384,0.04273600048489041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,10240,0.014094221923086377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,10240,0.013807999591032663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,7168,0.012103111379676394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,12288,0.034147553973727755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,8192,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,8192,0.012836444709036084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,6144,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,7168,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,7168,0.011387555963463254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,5120,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,8192,0.0265608893500434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,6144,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,6144,0.010759111079904767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,7168,0.02403644389576382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,4096,0.008370666868156856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,5120,0.00907733374171787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,6144,0.022647110952271357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,3584,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,4096,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,5120,0.0199946661790212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,4096,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,3072,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,4096,0.017883555756674874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,3584,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,2560,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,3072,0.007836444510353936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,3072,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,3584,0.016495111915800307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,2048,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,2560,0.0075635554061995606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,2560,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,10240,0.029892444610595703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,3072,0.014881778094503613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,1536,0.004402666870090696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,2048,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,2048,0.006302222195598815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,2560,0.013818666338920593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,1024,0.0032808888289663526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,1536,0.006017777654859755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,1536,0.006572444405820634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,768,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,2048,0.013078221844302284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,1536,0.011751111182901593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,1024,0.00565155554148886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,512,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,768,0.006031999985376994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,768,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,1024,0.011146666275130378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,256,0.00221422231859631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,3584,0.00776711106300354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,768,0.01035199976629681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,512,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,128,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,256,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,256,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,512,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,64,0.0019244444039132858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,128,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,256,0.009681777821646797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,32,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,65536,0.06051911248101128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,64,0.004972444640265571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,128,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,32,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,5120,0.009703111317422655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,51200,0.0491795539855957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,1024,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,65536,0.060376001728905566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,65536,0.057856890890333385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,16384,0.020879111356205408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,51200,0.04726755619049072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,51200,0.04623466730117798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,65536,0.13770222663879395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,512,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,12288,0.016604445046848722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,16384,0.0188426673412323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,16384,0.018555555078718398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,51200,0.11040711402893066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,10240,0.014454222387737699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,12288,0.015265777707099915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,12288,0.015294222368134392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,16384,0.04199644592073229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,8192,0.01277599980433782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,128,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,10240,0.01348355578051673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,10240,0.013531555732091268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,12288,0.03377688924471537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,8192,0.011633777783976661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,10240,0.029909332593282063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,8192,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,6144,0.010608000059922537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,7168,0.01053955571519004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,7168,0.011057777537239922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,5120,0.009311999711725447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,6144,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,7168,0.023768888579474554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,6144,0.01000355515215132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,4096,0.008055999875068665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,5120,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,6144,0.02199466692076789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,4096,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,3584,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,5120,0.01920088960064782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,4096,0.017145777742067974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,3584,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,7168,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,3072,0.006688888702127669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,3584,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,3584,0.016210666961140104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,2560,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,3072,0.007632888853549957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,3072,0.007240888973077138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,8192,0.026184888349639043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,2048,0.004975999808973736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,2560,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,2560,0.006689777804745569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,5120,0.00904888908068339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,2048,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,1536,0.003896000070704354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,2560,0.013795555465751223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,1024,0.0033208889265855155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,4096,0.008294222255547842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,1536,0.006650666809744305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,2048,0.013451555536852943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,1536,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,768,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,1024,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,1536,0.011752888560295105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,512,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,768,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,768,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,1024,0.011013333168294696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,256,0.0021582223061058256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,512,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,768,0.010863111250930362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,512,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,128,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,256,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,256,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,512,0.009684444301658207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,3072,0.015364444918102689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,64,0.0018008889423476325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,32,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,128,0.004960000101062986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,128,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,256,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,2048,0.0063528890411059065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,65536,0.05016266637378269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,64,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,51200,0.04064888755480448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,32,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,65536,0.05639288822809855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,65536,0.057816889550950795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,16384,0.017218665944205392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,1024,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,51200,0.04524533285035027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,51200,0.04527466826968723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,12288,0.013824889229403602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,65536,0.13754399617513022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,16384,0.01795288920402527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,10240,0.01386222243309021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,51200,0.10987377829021877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,12288,0.014190221826235453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,12288,0.01477599971824222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,16384,0.04199555516242981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,8192,0.01184800018866857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,10240,0.012445333103338877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,12288,0.03389244609408908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,10240,0.01311822235584259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,7168,0.010579555398888059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,8192,0.011758222348160215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,8192,0.011020444333553314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,10240,0.029473778274324205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,6144,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,8192,0.02615822189384037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,7168,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,7168,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,128,0.007626666790909237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,5120,0.01184177812602785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,6144,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,6144,0.010044444766309526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,4096,0.00963466697269016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,5120,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,5120,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,6144,0.02183288832505544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,16384,0.018586667047606576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,3584,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,4096,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,4096,0.007794666621420119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,3072,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,3584,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,4096,0.01699911057949066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,3584,0.016213332613309223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,3072,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,3072,0.014879110786649915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,2560,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,2560,0.007251555720965068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,2048,0.005977777971161737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,2560,0.013802666631009845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,5120,0.01923022170861562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,3584,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,2048,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,1536,0.003917333152559069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,2048,0.012612444659074148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,3072,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,2560,0.00675022229552269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,1024,0.0032622222271230486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,1536,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,1536,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,768,0.002965333354141977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,1536,0.012279999752839407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,1024,0.006298666612969504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,1024,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,512,0.002711999954448806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,1024,0.010687111152542962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,7168,0.023934221929974023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,256,0.002467555511328909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,512,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,768,0.010718222293588849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,2048,0.006527110934257507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,128,0.0021271111650599376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,512,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,256,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,128,0.004871110949251386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,64,0.0021137777302000257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,128,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,32,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,128,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,65536,0.03868177864286635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,64,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,32,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,768,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,51200,0.030838221311569214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,65536,0.05568710962931315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,65536,0.055192000336117215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,512,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,16384,0.016148444679048326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,65536,0.13743555545806885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,256,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,51200,0.044567111465666026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,51200,0.04373866650793287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,12288,0.013396444420019785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,16384,0.017867555220921833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,16384,0.016751110553741455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,10240,0.012068444656001197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,51200,0.1095306674639384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,12288,0.014053333136770459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,12288,0.014408000641398959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,8192,0.010787555740939246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,16384,0.04165244433614943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,12288,0.03379377722740173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,10240,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,10240,0.01278222186697854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,7168,0.010016000105275048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,8192,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,10240,0.02971733278698391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,768,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,6144,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,7168,0.01000533335738712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,7168,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,8192,0.025027554896142747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,5120,0.008741333252853816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,6144,0.009398221969604492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,7168,0.02331466641690996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,4096,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,6144,0.009369778136412302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,5120,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,5120,0.008609777523411645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,256,0.005663110978073544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,6144,0.020977778567208186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,4096,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,4096,0.007673777639865875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,5120,0.018583110637134977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,3072,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,4096,0.016864889197879367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,3584,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,3584,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,2560,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,3072,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,3584,0.015846222639083862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,8192,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,2560,0.006647111227114995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,2048,0.00452888881166776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,2560,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,3072,0.014828445182906257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,1536,0.0038515557017591265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,2048,0.0069208890199661255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,2560,0.014137778017255994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,2048,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,1024,0.003157333367400699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,1536,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,1536,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,2048,0.012721777790122561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,768,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,3584,0.0068711112770769335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,1024,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,1024,0.005738666488064661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,1536,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,512,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,768,0.005997333261701796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,768,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,1024,0.010747555229398938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,256,0.0021688888470331826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,512,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,512,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,3072,0.007135999699433644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,128,0.001869333287080129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,256,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,512,0.009688888986905416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,256,0.00508711114525795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,64,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,32,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,128,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,128,0.005542222410440445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,256,0.009328888522254096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,65536,0.02730488942729102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,64,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,32,0.00519822206762102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,51200,0.021943110558721755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,65536,0.05482044484880236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,65536,0.05283288823233711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,16384,0.012425777812798819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,51200,0.04401422209209866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,51200,0.042140444119771324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,65536,0.13669599427117243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,12288,0.011013333168294696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,16384,0.01686044368478987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,51200,0.10924800237019856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,10240,0.010064000056849586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,12288,0.013452444639470843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,768,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,16384,0.04150044586923387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,8192,0.011767110890812345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,12288,0.033024887243906655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,10240,0.012096000214417776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,10240,0.011752888560295105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,7168,0.010793777803579966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,8192,0.010776888993051318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,8192,0.010679999987284342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,128,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,10240,0.029149333635965984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,6144,0.009719111025333405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,7168,0.010052444206343757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,8192,0.024842666255103216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,5120,0.008616888688670265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,6144,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,6144,0.009114666945404476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,7168,0.023032888770103455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,4096,0.008172444171375698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,5120,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,6144,0.021006221572558086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,5120,0.008439111212889353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,16384,0.016202666693263583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,3584,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,4096,0.008295999632941352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,12288,0.013353777428468069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,5120,0.018897778458065458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,3072,0.0058497778243488735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,3584,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,3584,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,2560,0.005056000004212062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,4096,0.016851555969980028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,3072,0.007639110916190677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,3584,0.01571555601225959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,3072,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,2048,0.005538666827811136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,2560,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,2560,0.006783999916579988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,3072,0.014499555031458536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,1536,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,7168,0.010116444693671333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,2048,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,2048,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,2560,0.013461333182122974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,1024,0.003726222034957674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,2048,0.013052444491121503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,1536,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,768,0.003411555455790626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,1024,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,1024,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,1536,0.011696889168686338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,512,0.0029315555261241067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,1024,0.010773333410422007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,768,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,4096,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,768,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,256,0.002493333278430833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,512,0.0058782220714622075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,512,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,768,0.010363555616802638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,128,0.0021928888228204516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,256,0.005624889085690181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,256,0.005065777649482091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,64,0.002143111079931259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,512,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,32,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,128,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,128,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,256,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,65536,0.02644177774588267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,64,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,128,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,32,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,51200,0.022630222969584998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,65536,0.054438223441441856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,1536,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,65536,0.052003555827670626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,51200,0.043515556388431124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,51200,0.04156088829040527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,12288,0.009385777844323052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,16384,0.016872889465755887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,16384,0.016146666473812528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,51200,0.10877688725789388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,10240,0.008653333617581261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,12288,0.013420444395807056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,16384,0.04111466805140177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,12288,0.013074666261672974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,8192,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,10240,0.012074666718641916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,12288,0.03258488906754388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,10240,0.011713777979214987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,7168,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,8192,0.01035377797153261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,8192,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,10240,0.028912888632880315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,6144,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,7168,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,7168,0.00976533359951443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,8192,0.024728000164031982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,5120,0.007104000283612146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,6144,0.009471110999584198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,6144,0.0090524446633127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,7168,0.022358222140206233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,4096,0.006599111275540457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,65536,0.13676622178819445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,5120,0.00868977771864997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,6144,0.020908445119857788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,3584,0.00630044440428416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,4096,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,5120,0.018567999203999836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,4096,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,3072,0.0057813334796163775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,3584,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,4096,0.01648266613483429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,2560,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,3072,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,3072,0.0148053334818946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,2560,0.006653333289755716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,2048,0.004523555437723796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,2560,0.007297777467303806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,16384,0.010936000280910067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,2560,0.01346666696998808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,1536,0.0038613333470291565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,2048,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,2048,0.006908444480763541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,5120,0.008608000146018134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,1536,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,2048,0.01275288893116845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,1536,0.011389333340856763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,3584,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,1024,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,768,0.0028213332924577924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,1024,0.00619911112719112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,3072,0.007632888853549957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,512,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,768,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,768,0.005528000079923206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,512,0.005299555758635203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,512,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,256,0.0021253333737452826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,512,0.009725333087974125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,256,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,1024,0.0031537777847713898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,1536,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,128,0.001793777777089013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,256,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,256,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,128,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,128,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,32,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,3584,0.015847111741701763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,1024,0.011047110789351992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,65536,0.02085599965519375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,64,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,32,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,51200,0.01756088932355245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,768,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,65536,0.05091466506322225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,51200,0.04120533333884345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,65536,0.13627999358707005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,16384,0.01145066652033064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,51200,0.10929333501391941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,16384,0.01586933268441094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,12288,0.009406222237481011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,16384,0.015817777978049386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,768,64,0.001793777777089013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,128,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,12288,0.01295199990272522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,12288,0.013345777988433838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,10240,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,10240,0.012070222033394707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,8192,0.00776711106300354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,65536,0.05345511105325487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,12288,0.03282311227586534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,10240,0.028151111470328435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,8192,0.01070755554570092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,7168,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,51200,0.040904889504114784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,8192,0.01032977799574534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,6144,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,7168,0.01000533335738712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,8192,0.024659555819299486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,7168,0.009712888962692684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,5120,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,6144,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,6144,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,4096,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,7168,0.022272888157102797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,16384,0.04079644547568427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,5120,0.008734222087595198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,5120,0.00831911133395301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,3584,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,6144,0.02030666669209798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,4096,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,3072,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,5120,0.019142222073343065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,3584,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,3584,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,2560,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,10240,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,3072,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,3072,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,3584,0.01575644479857551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,2048,0.004413333204057482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,2560,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,3072,0.015115555789735583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,1536,0.0038275557259718576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,2048,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,2560,0.013106666505336761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,2048,0.012392889294359418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,1536,0.005990222096443176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,1024,0.003156444471743372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,1536,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,1024,0.006116444451941385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,1536,0.011710222396585675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,768,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,1024,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,4096,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,512,0.002536000063021978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,1024,0.010678222609890832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,768,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,768,0.005535999933878581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,4096,0.016564443707466125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,256,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,512,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,768,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,512,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,128,0.0020071110791630214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,256,0.00563733321097162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,256,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,64,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,512,0.009391110804345872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,512,32,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,2048,0.0067528887755341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,128,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,256,0.009167999857001835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,128,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,65536,0.014742222097184924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,64,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,128,0.00701333334048589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,51200,0.013086222112178802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,65536,0.047255999512142606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,32,0.004866666677925321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,16384,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,65536,0.049230221245023936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,51200,0.037736889388826154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,51200,0.039367109537124634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,65536,0.13588178157806396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,12288,0.006800888727108638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,16384,0.015631111131774057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,51200,0.10806577735477024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,12288,0.013296888934241401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,12288,0.013019555144839816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,2560,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,10240,0.011561777856614856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,8192,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,10240,0.012054222325483957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,10240,0.028201778729756672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,8192,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,8192,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,8192,0.02502222193611993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,7168,0.009644444617960189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,16384,0.015808888607554965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,6144,0.006086222413513396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,7168,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,10240,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,7168,0.02273688879277971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,6144,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,6144,0.009316444396972656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,5120,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,12288,0.0322382218307919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,4096,0.006004444426960415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,5120,0.008987555901209513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,6144,0.020592888196309406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,5120,0.008336000144481659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,3584,0.00565422202150027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,4096,0.007380444142553542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,4096,0.008039111064540016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,5120,0.0185208883550432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,7168,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,3072,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,3584,0.00776800016562144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,3584,0.007297777467303806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,4096,0.016881777180565726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,2560,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,3072,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,3072,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,2560,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,3584,0.015503111812803479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,2560,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,3072,0.01478488908873664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,1536,0.003576888806290097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,2048,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,2048,0.006247111078765657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,2560,0.013736888766288757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,1024,0.0031866667171319327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,1536,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,2048,0.013038222160604266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,768,0.0028035555862718155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,1024,0.006146666490369373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,1024,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,1536,0.012068444656001197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,512,0.0024986666523747975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,768,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,768,0.005483555710977978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,512,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,512,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,256,0.002116444417172008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,768,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,2048,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,128,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,256,0.005056888692908817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,16384,0.04051911168628269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,64,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,256,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,128,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,128,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,256,32,0.0017893332988023758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,65536,0.011125333607196808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,1536,0.0058666666348775225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,64,0.00499288903342353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,51200,0.010264000131024255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,32,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,65536,0.047925333182017006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,65536,0.04653155472543505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,16384,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,1024,0.010683555569913654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,51200,0.03764177693261041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,51200,0.038375112745496966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,65536,0.13509778181711832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,12288,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,16384,0.015826667348543804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,16384,0.015533333023389181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,10240,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,51200,0.10759377479553223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,12288,0.0127697777416971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,12288,0.013390222357379066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,512,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,16384,0.04003555576006571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,8192,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,10240,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,10240,0.011375111010339526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,7168,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,8192,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,10240,0.028782222006056044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,128,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,8192,0.02458933326933119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,6144,0.00518222235971027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,7168,0.009536888864305284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,6144,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,5120,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,7168,0.022341334157519873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,6144,0.020251555575264823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,5120,0.008173333273993598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,5120,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,5120,0.01832444469134013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,12288,0.03228088882234361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,4096,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,3584,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,8192,0.010122666756312052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,4096,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,4096,0.016531555189026725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,3584,0.007772444850868649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,7168,0.010021333065297868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,3072,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,3072,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,2560,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,6144,0.009004444711738164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,3072,0.007002666592597961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,3072,0.01479022204875946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,2048,0.004646222210592694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,2560,0.00630666646692488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,4096,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,2560,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,1536,0.003530666646030214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,2048,0.006770666688680649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,2048,0.006308444258239534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,1536,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,2048,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,1024,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,1536,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,768,0.0028497777465316984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,1024,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,1536,0.011553777588738335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,1024,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,512,0.002524444419476721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,768,0.005304000029961268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,768,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,3584,0.015459555718633862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,256,0.0021262222694026097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,512,0.005843555761708154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,512,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,2560,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,128,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,256,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,256,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,128,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,128,0.005035555611054103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,256,0.008727110922336578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,64,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,128,32,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,65536,0.008580444587601556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,128,0.007252444823582967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,51200,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,32,0.004607111215591431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,16384,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,3584,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,65536,0.04686933424737719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,12288,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,10240,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,51200,0.037754668129815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,1024,0.010724444356229572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,8192,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,16384,0.015508444772826301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,7168,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,12288,0.012506666282812754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,6144,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,5120,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,10240,0.011438222395049201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,768,0.010676444404655032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,4096,0.004535111288229625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,3584,0.004572444491916233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,7168,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,3072,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,6144,0.00868977771864997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,5120,0.008112000094519721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,2560,0.004566222015354368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,2048,0.004237333519591225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,4096,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,1536,0.0035235554807715942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,3584,0.007104000283612146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,1024,0.0028275555620590844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,3072,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,768,0.002537777854336633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,2560,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,2048,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,512,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,64,0.005139555368158552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,1536,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,256,0.0021582223061058256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,128,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,768,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,64,0.0018275556051068837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,512,0.004958222309748332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,64,32,0.001782222133543756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,256,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,128,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,64,0.004567999806669024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,51200,0.008556444611814287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,32,0.004530666602982415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,16384,0.005695999910434087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,8192,0.010092444717884064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,65536,0.047213332520590894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,512,0.009762666291660732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,12288,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,10240,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,8192,0.00519822206762102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,51200,0.03770577907562256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,7168,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,16384,0.015466666883892484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,12288,0.012801777985360889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,6144,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,10240,0.011446221835083432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,5120,0.004588444613748127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,8192,0.010335110955768162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,7168,0.009451555709044138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,3584,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,1024,0.00553688903649648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,6144,0.008980444735950893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,3072,0.004505777938498391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,5120,0.00832088871134652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,2560,0.004577777865860197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,2048,0.004514666481150521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,3584,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,65536,0.01036800030204985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,1536,0.0035004446076022256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,2560,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,1024,0.0028408887899584244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,768,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,2048,0.006230222268237009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,512,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,1536,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,1024,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,256,0.002159111201763153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,128,0.001818666648533609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,768,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,64,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,256,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,32,0.0018586667461527719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,128,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,96,32,4096,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,32,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,4096,0.007385777930418651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,3072,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,16384,0.37297601170010036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,16384,0.20784444279140898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,51200,0.6031440099080404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,512,0.005088888936572605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,12288,0.277091556125217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,16384,0.13299466503991023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,51200,1.1407830980088975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,10240,0.20679110950893828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,12288,0.1485875580045912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,64,0.00462666650613149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,12288,0.10161777999666001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,8192,0.17238488462236193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,7168,0.14986133575439453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,10240,0.08393333355585735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,8192,0.0685911112361484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,8192,0.10653689172532822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,7168,0.0927928884824117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,10240,0.12569422192043728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,51200,0.40589067671034074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,5120,0.10882933272255792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,4096,0.0919075542026096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,6144,0.05367111166318258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,6144,0.08245333035786946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,5120,0.04596799943182203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,7168,0.06950044631958008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,5120,0.06613600254058838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,3072,0.06814222203360663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,4096,0.054382221566306226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,4096,0.039204445150163435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,3584,0.03472800056139628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,2560,0.05723288986417982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,3584,0.04851466748449537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,3072,0.030439999368455675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,3072,0.04238399863243103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,2560,0.027216888136333887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,1536,0.03659644391801622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,2048,0.024988444315062627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,2560,0.036927110619015165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,1024,0.027173333697848853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,2048,0.03182755576239692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,768,0.020089776979552377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,1536,0.018703111343913607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,6144,0.12820088863372803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,1024,0.014953777194023132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,512,0.013793778088357715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,1024,0.02976800004641215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,768,0.013759111364682516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,256,0.008440888590282863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,768,0.021897777915000916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,512,0.012303111453851065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,128,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,512,0.016706667012638517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,256,0.011033777561452655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,3584,0.07812266879611544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,64,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,128,0.011038222246699862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,32,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,128,0.011414222419261932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,2048,0.04689155684577095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,256,0.012431110772821637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,64,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,32,0.015204444527626038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,1536,0.04128355450100369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,16384,0.2780391110314263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,51200,0.47535377078586155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,51200,0.8569635815090604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,65536,1.090695063273112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,12288,0.2093964417775472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,16384,0.10700888766182794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,16384,0.15933244758182102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,51200,0.33563910590277773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,65536,0.41838667127821183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,10240,0.17590310838487414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,12288,0.11804178025987412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,8192,0.14369155300988093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,12288,0.0848311119609409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,8192,0.05728089147143894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,7168,0.12614577346377903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,65536,0.6275715298122829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,6144,0.10849155320061578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,8192,0.0849999984105428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,7168,0.05051022105746799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,6144,0.04411288764741686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,7168,0.07329422235488892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,5120,0.09171555439631145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,6144,0.063028441535102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,4096,0.07398577531178792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,3584,0.06600177950329252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,5120,0.053463111321131386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,4096,0.0314684444003635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,3072,0.05780711438920763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,4096,0.04409511221779717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,3584,0.039735111925337054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,3584,0.028586665789286297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,2560,0.048615111245049365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,3072,0.026142223013771906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,2048,0.03990755478541056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,3072,0.03477155499988132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,1536,0.03154755632082621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,10240,0.10026311212115818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,2560,0.030402666992611353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,1024,0.023223999473783705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,2560,0.022720888257026672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,2048,0.026536888546413843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,1536,0.016248888439602323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,768,0.016659556163681876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,1536,0.03325422273741828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,512,0.011971555650234222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,1024,0.013282666603724161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,768,0.01219377749496036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,1024,0.0236453331179089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,256,0.00793333351612091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,768,0.018522666560279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,512,0.010848888920413123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,128,0.005487999982304043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,512,0.01440711153878106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,64,0.0041911109454101985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,256,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,5120,0.038395557138654925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,32,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,128,0.009937778115272522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,64,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,32,0.013090666797426013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,65536,0.3320017655690511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,10240,0.07732266849941678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,65536,0.2254577742682563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,2048,0.021941334009170532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,16384,0.08694044748942058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,65536,0.15420533551110163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,51200,0.11706044938829209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,12288,0.06649600134955512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,16384,0.04382044408056471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,10240,0.0548257761531406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,256,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,12288,0.03228266702757941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,128,0.00976533359951443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,12288,0.04575555523236593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,8192,0.047203554047478564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,10240,0.03089600139194065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,7168,0.03985688752598233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,10240,0.03936533464325799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,8192,0.023523555861579046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,6144,0.03455288873778449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,7168,0.02578488820128971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,8192,0.03380089004834493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,7168,0.02951822347111172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,5120,0.02938577863905165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,6144,0.02623555560906728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,6144,0.01978755493958791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,4096,0.02443644404411316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,5120,0.017498667041460674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,5120,0.02290222214327918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,4096,0.014850666125615438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,51200,0.17193955845303008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,4096,0.01921333372592926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,3072,0.019094222121768527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,16384,0.06063377857208252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,3584,0.013955555028385587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,2560,0.01583377851380242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,3584,0.017343110508388944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,3072,0.013112000293201871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,2048,0.013088888592190213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,3072,0.015505777464972602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,1536,0.010327999790509542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,2560,0.013820444544156393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,2048,0.012208888928095499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,1024,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,2048,0.012663111090660095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,1536,0.008929777476522658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,768,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,1536,0.0140657772620519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,1024,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,1024,0.011296889020336999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,512,0.005192000004980299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,768,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,256,0.0035093331502543557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,768,0.009399111072222391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,512,0.0070168889231152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,128,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,3584,0.021566222111384075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,512,0.007631111476156447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,256,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,256,0.006436444405052397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,128,0.006664888726340399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,32,0.0028311111446883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,128,0.006309333360857434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,64,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,2560,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,32,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,65536,0.2724844349755181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,65536,0.1724720001220703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,51200,0.21756533781687418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,16384,0.06934133503172132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,65536,0.12683110766940647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,51200,0.1316266722149319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,12288,0.05489600035879347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,10240,0.0455075568623013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,16384,0.03752177622583177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,51200,0.10390488968955146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,51200,0.2585555447472466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,16384,0.0469688872496287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,8192,0.038450668255488075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,12288,0.03627466824319627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,7168,0.033026665449142456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,64,0.002528888897763358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,10240,0.02572533322705163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,10240,0.03155644403563605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,6144,0.028498666154013738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,5120,0.024657777614063684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,7168,0.022280888424979314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,8192,0.027415111660957336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,7168,0.023916444844669763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,4096,0.021141333712471858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,6144,0.021190222766664293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,5120,0.016121778223249648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,3584,0.018225777480337355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,4096,0.014655111564530266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,5120,0.017948443690935772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,3072,0.015942222542232938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,4096,0.01537333263291253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,2560,0.013384000294738345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,3584,0.013501333693663279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,3584,0.014308444327778287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,2048,0.011398221883508893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,3072,0.01309866706530253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,2560,0.011568889021873474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,1536,0.009510222408506606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,2560,0.011808888779746162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,2048,0.010796444283591377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,1024,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,2048,0.011016888750924004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,12288,0.030010667112138536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,1536,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,768,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,1536,0.012286221815480126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,512,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,1024,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,8192,0.021907554732428655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,1024,0.009496000077989366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,256,0.0035146665241983202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,768,0.007979555262459649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,128,0.003003555453485913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,6144,0.017596445149845548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,512,0.006726222319735422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,256,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,256,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,64,0.00249422217408816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,128,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,32,0.002537777854336633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,128,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,3072,0.0127368892232577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,64,0.005970666805903117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,65536,0.21302221881018746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,32,0.006996444529957241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,51200,0.1687715583377414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,65536,0.1649057732688056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,16384,0.05975288814968533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,65536,0.12356711758507623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,51200,0.09960444768269856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,51200,0.1266684399710761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,12288,0.04628088739183214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,10240,0.038042667839262224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,16384,0.03571910990609063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,768,0.007376000285148621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,16384,0.04498222139146593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,512,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,12288,0.03466755482885573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,10240,0.02590222160021464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,7168,0.02836977773242527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,10240,0.030200888713200886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,8192,0.021689777572949726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,6144,0.0244159996509552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,8192,0.025961778230137292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,5120,0.020834666159417894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,7168,0.022597332795461018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,6144,0.01755466726091173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,4096,0.017687110437287223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,6144,0.019948444432682462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,5120,0.015791111522250705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,5120,0.01738666660255856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,3584,0.015184000134468079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,4096,0.01385600037044949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,3072,0.01346933344999949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,4096,0.01511111193233066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,3584,0.01308444473478529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,2560,0.011519999967681037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,3584,0.013896889156765409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,3072,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,2048,0.009631111390060848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,3072,0.012823999755912356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,12288,0.028042667441897925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,2560,0.010789333118332757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,2048,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,1536,0.008298666940795051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,1024,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,2048,0.010617777705192566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,1536,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,768,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,1536,0.010831111007266574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,7168,0.021125333176718816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,512,0.00388711111413108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,1024,0.008632889224423302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,256,0.003181333343187968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,768,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,512,0.006801777829726537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,128,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,512,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,256,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,256,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,64,0.00253866674999396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,32,0.002543111021320025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,128,0.005985777825117111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,64,0.005956444475385878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,32,0.006654222392373615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,8192,0.03267733256022135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,2560,0.011689778003427716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,51200,0.1327715582317776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,65536,0.12228355142805312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,65536,0.07056444221072726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,1024,0.007620444728268518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,16384,0.046966221597459584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,51200,0.09359111388524373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,12288,0.036347554789649114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,16384,0.02312711046801673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,768,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,51200,0.05727110968695747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,10240,0.030143999391131934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,16384,0.034176889393064715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,12288,0.026711111267407734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,10240,0.016598222984208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,128,0.0063279999627007385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,7168,0.022260444031821355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,8192,0.014429334137174817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,6144,0.01943644384543101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,8192,0.019952888290087383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,7168,0.013743999931547375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,5120,0.016640888320075143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,65536,0.1684622234768338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,7168,0.017493334081437852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,4096,0.01389955480893453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,6144,0.015799111790127225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,5120,0.011515555282433828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,3584,0.012395555774370829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,4096,0.010574222438865237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,5120,0.014133334159851074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,4096,0.012542222109105853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,3072,0.011054221954610614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,3584,0.009587555295891231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,3584,0.011707555916574268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,2560,0.009659555223253038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,3072,0.008775999976529015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,3072,0.010781333678298526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,12288,0.018577777677112155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,8192,0.026100445124838088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,2560,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,1536,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,2048,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,2048,0.009007111191749573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,1024,0.005314666777849197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,1536,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,1536,0.009269333548016017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,768,0.004171555654870139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,1024,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,1024,0.0075599998235702515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,512,0.003181333343187968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,768,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,768,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,256,0.0028213332924577924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,512,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,512,0.006054222169849608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,128,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,256,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,256,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,64,0.0024746666765875286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,32,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,128,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,128,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,10240,0.02334311107794444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,65536,0.16092356046040854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,64,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,32,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,6144,0.0123848890264829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,51200,0.12839644485049778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,16384,0.042753779225879245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,65536,0.11894933382670085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,12288,0.03362755642996894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,51200,0.05498488744099935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,51200,0.09104622072643703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,16384,0.021651556094487507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,10240,0.028120001157124836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,16384,0.03338755501641168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,8192,0.0237128883600235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,12288,0.017895110779338412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,12288,0.025891555680169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,10240,0.015791111522250705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,7168,0.020652444826232064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,10240,0.022456000248591106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,8192,0.014112000664075216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,6144,0.017902221944597032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,2048,0.008267555799749162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,7168,0.013032889200581444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,5120,0.015445333388116626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,7168,0.01719377769364251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,6144,0.012068444656001197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,4096,0.013406222065289816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,2560,0.009961778091059791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,6144,0.015783111254374187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,5120,0.01110400011142095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,3584,0.0118568887313207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,3072,0.010312889185216691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,5120,0.013811555173661975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,4096,0.009848888549539778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,2560,0.008997333546479544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,3584,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,4096,0.012637333737479316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,2048,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,3072,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,65536,0.06804177496168348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,1536,0.00655644428398874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,2560,0.007720888488822513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,3072,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,1024,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,2560,0.009917333722114563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,2048,0.007440889047251807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,768,0.004015999949640698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,1536,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,2048,0.0083137775460879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,512,0.0031973332580592898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,1024,0.006554666492674086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,1536,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,256,0.0028239999794297745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,1024,0.007600000335110559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,768,0.006298666612969504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,128,0.0025351111673646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,512,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,768,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,64,0.002479111154874166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,256,0.005672889037264719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,128,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,32,0.002504000026318762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,8192,0.01926933394538032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,128,0.0053386667536364655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,65536,0.14101066854265001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,64,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,32,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,51200,0.11091111103693645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,65536,0.11214844385782878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,65536,0.06635110908084445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,16384,0.040856887896855675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,51200,0.08616177903281318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,51200,0.05333066648907132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,16384,0.020793777373101976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,16384,0.03221777743763394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,12288,0.03176800078815884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,10240,0.02773688899146186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,12288,0.017183999220530193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,8192,0.021394666698243883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,10240,0.015457777513398064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,10240,0.021312889125612047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,512,0.005974222388532426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,7168,0.019363555643293593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,8192,0.013713777893119387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,256,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,6144,0.017436444759368896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,7168,0.01278222186697854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,5120,0.014003554979960123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,6144,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,7168,0.016715556383132935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,6144,0.015365334020720588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,4096,0.013062222136391534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,5120,0.01071200023094813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,3584,0.011585777832402123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,4096,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,5120,0.013474666410022311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,3072,0.009683555199040307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,4096,0.01201333353916804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,3584,0.008619555168681675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,2560,0.0090506664580769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,3584,0.011352889239788055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,3072,0.007923555870850882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,2048,0.007770666645632849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,3072,0.01039644413524204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,2560,0.007668444679843054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,12288,0.024949333733982507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,2560,0.009422221945391761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,2048,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,1024,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,2048,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,1536,0.006913777854707506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,768,0.004197333421972063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,8192,0.018713777263959248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,1536,0.008020444048775567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,512,0.0031991110493739447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,1024,0.0071928890215026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,768,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,256,0.002564444517095884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,768,0.006341333190600078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,512,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,512,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,3584,0.011527111132939657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,256,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,64,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,256,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,32,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,128,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,128,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,65536,0.11140178309546576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,64,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,32,0.006199999815887875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,51200,0.08691644668579102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,16384,0.03290044599109226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,1536,0.0064088888466358185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,65536,0.06339110930760701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,12288,0.027217777238951787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,51200,0.05062222149637011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,51200,0.08386577500237359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,10240,0.023968888653649226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,1024,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,16384,0.030660443835788306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,8192,0.02146311104297638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,12288,0.02369955513212416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,10240,0.015047111445003085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,7168,0.018526222970750596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,128,0.002451555596457587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,8192,0.013275555438465543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,6144,0.013464888764752282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,8192,0.018210666047202218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,7168,0.012494222157531314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,5120,0.012034667034943899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,7168,0.016421332955360413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,6144,0.011289777855078379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,4096,0.010461333725187512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,6144,0.01499644418557485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,5120,0.010016000105275048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,3584,0.009322666459613377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,5120,0.013286222186353473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,4096,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,3072,0.009177777502271865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,65536,0.10718844334284465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,4096,0.011839110818174152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,3584,0.008634666601816813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,2560,0.008294222255547842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,3584,0.01109155515829722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,16384,0.020234665936893888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,2048,0.007520000139872233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,3072,0.009764444496896531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,2560,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,12288,0.016661332713233102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,1536,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,2560,0.008614222208658854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,1024,0.004597333156400257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,2048,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,1536,0.006646222124497096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,2048,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,768,0.003909333298603694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,10240,0.020807110601001315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,1536,0.0076248885856734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,512,0.0035235554807715942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,1024,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,256,0.0028702221396896574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,1024,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,768,0.006091555373536215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,128,0.0023235556566052968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,512,0.005976888868543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,768,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,256,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,512,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,32,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,128,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,128,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,65536,0.0875928865538703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,64,0.005403555515739653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,32,0.005956444475385878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,51200,0.07125066386328803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,65536,0.06093511316511366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,65536,0.06551822026570638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,16384,0.029000888268152874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,51200,0.0519644452465905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,51200,0.04892622099982368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,12288,0.023034666975339253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,16384,0.020167110694779288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,10240,0.02015911042690277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,12288,0.01612977849112617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,12288,0.01612444387541877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,8192,0.01534222231970893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,10240,0.014473777678277759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,64,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,10240,0.014447111222479077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,256,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,8192,0.012960000170601739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,8192,0.013066666821638743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,6144,0.011808888779746162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,7168,0.011753777662913004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,5120,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,7168,0.011902222202883827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,6144,0.010689777632554373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,4096,0.00906044410334693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,6144,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,3072,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,5120,0.009344889058007134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,3584,0.008084444536103142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,5120,0.010032888915803697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,4096,0.008623110751310984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,3072,0.007359999749395583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,4096,0.008987555901209513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,3584,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,2560,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,16384,0.019596444235907663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,2048,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,3584,0.008319999608728621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,3072,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,1536,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,2560,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,3072,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,2048,0.006990222467316522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,1024,0.0034844444857703317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,2560,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,768,0.0031600000543726813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,1536,0.006656000183688269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,2048,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,512,0.0028328889360030494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,1024,0.006409777949253718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,1536,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,256,0.002495999965402815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,768,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,7168,0.013378666506873237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,1024,0.005970666805903117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,512,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,768,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,64,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,256,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,32,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,512,0.005392888767851724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,128,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,256,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,65536,0.07658844523959689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,128,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,64,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,51200,0.06191733148362902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,32,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,16384,0.026782222919993933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,65536,0.0642755561404758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,65536,0.06021333403057522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,12288,0.021260443660948012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,51200,0.04805066519313388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,51200,0.05011555552482605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,16384,0.019276445110638935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,10240,0.018548443913459778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,8192,0.01681777834892273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,12288,0.016011555989583332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,12288,0.01549600064754486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,7168,0.012828444441159567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,10240,0.014137778017255994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,8192,0.01240355521440506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,6144,0.011331555744012197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,8192,0.012674666941165924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,7168,0.0110871113008923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,5120,0.00981955561372969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,7168,0.011365333365069495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,6144,0.010313777459992303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,4096,0.009161777794361115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,128,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,6144,0.010812444819344414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,5120,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,3584,0.007910221815109253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,5120,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,3072,0.0070222218831380205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,3584,0.007939555578761632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,3584,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,2560,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,3072,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,3072,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,2048,0.005295111073387994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,2560,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,16384,0.019116444720162284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,2560,0.006907555378145642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,1536,0.004426666845877965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,10240,0.014441778262456259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,2048,0.006945778098371293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,1024,0.0035128887328836652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,2048,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,768,0.0031840000301599503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,1024,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,1536,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,512,0.0028284444577164124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,1024,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,768,0.006122666514582104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,256,0.002493333278430833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,768,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,512,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,128,0.002208000048995018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,512,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,4096,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,64,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,4096,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,256,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,32,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,128,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,128,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,65536,0.06764533122380574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,64,0.005199111170238919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,32,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,65536,0.058766219351026744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,16384,0.02215733296341366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,51200,0.04674577713012695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,51200,0.04843022094832527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,1536,0.006728888799746831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,12288,0.018395556343926322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,16384,0.019208000765906442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,16384,0.018629334039158292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,10240,0.015326221783955893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,8192,0.013427555561065674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,12288,0.015860444969601102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,10240,0.013903111219406128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,12288,0.01515111161602868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,7168,0.012248000337017907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,8192,0.01145155562294854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,10240,0.013629333012633853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,6144,0.010948444406191507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,256,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,8192,0.012279999752839407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,5120,0.009654222263230218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,7168,0.01072533345884747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,4096,0.008610666626029545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,6144,0.009748444788985783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,7168,0.011353777514563667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,5120,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,6144,0.0105608892109659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,51200,0.05432266659206814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,5120,0.009299555586444007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,4096,0.008280889027648503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,3072,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,4096,0.008295999632941352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,2560,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,65536,0.06163644128375583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,3072,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,2048,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,2560,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,3072,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,1536,0.004209777961174647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,2560,0.006800888727108638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,1024,0.0034942221310403612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,1536,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,1536,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,1024,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,768,0.0029395555870400537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,1024,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,768,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,512,0.002500444443689452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,768,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,3584,0.0075839997993575195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,256,0.0023048888478014204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,512,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,3584,0.008058666355080074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,512,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,128,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,256,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,3584,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,64,0.0019111111760139465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,256,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,32,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,128,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,2048,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,128,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,51200,0.05168266759978401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,2048,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,32,0.005380444642570283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,65536,0.05754577451282077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,65536,0.05923555294672648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,51200,0.0457822216881646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,51200,0.04653333293067085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,12288,0.01641688909795549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,16384,0.018980445133315194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,10240,0.014140443669425117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,16384,0.018114666144053142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,12288,0.015146666102939181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,8192,0.012338666452301873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,12288,0.014831110835075378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,10240,0.013063111239009433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,7168,0.01166133334239324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,10240,0.013349333571063148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,8192,0.011712888876597086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,6144,0.010423111418883005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,8192,0.01203644441233741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,7168,0.01068800025516086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,5120,0.009328888522254096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,7168,0.010959111154079437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,6144,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,65536,0.06336266464657254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,4096,0.008206222620275285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,6144,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,5120,0.008967999782827165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,3584,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,64,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,5120,0.008966222405433655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,3072,0.006783999916579988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,4096,0.00830844458606508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,2560,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,16384,0.019896000623703003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,3584,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,4096,0.007939555578761632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,2048,0.00517511119445165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,3072,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,3584,0.0075351107451650835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,1536,0.0038720000949170855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,3072,0.007246221933099959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,2560,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,1024,0.0032986667421129015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,2048,0.006945778098371293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,2560,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,768,0.003132444495956103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,1536,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,2048,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,512,0.002546666603949335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,1536,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,256,0.002181333386235767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,768,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,1024,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,128,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,512,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,768,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,64,0.0018168888572189545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,256,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,32,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,512,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,128,0.005664888769388199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,65536,0.052129778597089976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,256,0.005112888912359874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,128,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,51200,0.0424195561144087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,64,0.005197333378924264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,32,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,65536,0.05612711111704508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,65536,0.057223108079698354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,51200,0.04482311010360718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,51200,0.04452266626887851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,16384,0.01846311158604092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,10240,0.012660444610648684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,16384,0.017854221993022494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,12288,0.014803555276658801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,8192,0.011634666886594562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,12288,0.014489778214030795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,7168,0.010426667001512317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,8192,0.011175110936164856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,8192,0.011360888679822287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,1024,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,6144,0.009672000176376766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,7168,0.010430222584141625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,6144,0.009639999932712978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,5120,0.009128000173303816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,6144,0.009322666459613377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,16384,0.01666844387849172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,4096,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,5120,0.008986666798591614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,12288,0.013455111119482251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,5120,0.008569777839713627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,3584,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,4096,0.00831911133395301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,3072,0.006175111151403851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,4096,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,3584,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,2560,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,10240,0.012450666891203986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,3584,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,3072,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,10240,0.013085333009560903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,3072,0.007280888656775157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,2560,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,1536,0.003916444463862313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,2560,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,2048,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,1024,0.0032462223122517266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,1536,0.006687110910813014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,2048,0.006313777632183499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,7168,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,768,0.003059555465976397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,512,0.002768888862596618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,1536,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,768,0.005971555494599872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,256,0.0024542222834295696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,512,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,768,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,128,0.0021537778278191886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,512,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,256,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,64,0.002128888956374592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,32,0.0021582223061058256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,128,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,256,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,128,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,65536,0.040299554665883384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,51200,0.03219733304447598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,64,0.004979555391603046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,32,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,65536,0.05534488956133524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,16384,0.01388177772363027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,2048,0.004575110971927643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,51200,0.044477333625157676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,12288,0.013167999684810638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,51200,0.042576889197031655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,16384,0.017623111605644226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,10240,0.012387555506494312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,12288,0.013788444300492605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,16384,0.017277333471510146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,8192,0.011006222003036074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,10240,0.012102222277058495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,12288,0.014204444156752693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,7168,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,1024,0.0063546668324205614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,10240,0.012381333443853589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,6144,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,8192,0.011001778145631155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,1024,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,5120,0.008760888543393876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,8192,0.010520888699425591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,7168,0.010258666343159145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,4096,0.00777599960565567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,6144,0.009657777845859528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,7168,0.009766221874290043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,3584,0.006335999816656113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,5120,0.008946666287051307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,6144,0.009319999979601966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,3072,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,5120,0.008661333057615492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,3584,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,2560,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,4096,0.007620444728268518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,2048,0.004495111190610462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,3072,0.0074826669361856245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,3584,0.007328888608349695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,1536,0.003863999827040566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,2560,0.007325333025720384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,2048,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,2560,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,65536,0.05446311169200473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,1024,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,768,0.0028124445428450904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,2048,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,1536,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,1024,0.006476444502671559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,512,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,768,0.006055999961164262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,256,0.002218666589922375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,1024,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,768,0.005531555662552516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,512,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,128,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,512,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,256,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,64,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,256,0.004952888935804367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,32,0.0018311111877361934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,128,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,128,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,65536,0.029509332444932725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,4096,0.008291555775536431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,64,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,51200,0.023935111032591924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,32,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,65536,0.054435554477903575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,16384,0.011422221859296164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,65536,0.05204800102445814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,3072,0.007017778025733099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,12288,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,51200,0.041251556740866765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,16384,0.016176000237464905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,10240,0.009687999884287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,16384,0.016895999511082966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,12288,0.013451555536852943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,8192,0.011649777491887411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,12288,0.01311644415060679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,10240,0.012082666986518435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,7168,0.011012444065676795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,10240,0.011772444678677453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,8192,0.01072177787621816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,6144,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,8192,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,7168,0.010012444522645738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,5120,0.00885333369175593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,6144,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,7168,0.009755555954244401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,4096,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,6144,0.00908799966176351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,5120,0.008737777670224508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,3584,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,5120,0.008364444805516137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,3072,0.005630222045713001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,3584,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,4096,0.007815111014578078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,2560,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,3584,0.007449777589903937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,3072,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,2048,0.005426666802830166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,3072,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,2560,0.007062222394678328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,51200,0.04407911168204414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,1536,0.0046533333758513136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,2560,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,1024,0.0037804444630940757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,2048,0.006865777903132969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,1536,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,1536,0.006526222245560751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,768,0.003220444545149803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,2048,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,512,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,1024,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,768,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,1024,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,256,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,768,0.005538666827811136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,128,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,512,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,256,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,64,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,32,0.0021671110557185277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,256,0.005208000126812193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,128,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,65536,0.023944889505704243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,128,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,4096,0.008006222546100616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,51200,0.020045333438449435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,64,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,32,0.00508711114525795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,65536,0.054478221469455294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,51200,0.04360266526540121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,65536,0.05076444480154249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,12288,0.009425777528021071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,51200,0.04077422287729051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,16384,0.016132444143295288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,10240,0.008691555923885768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,16384,0.015835555063353646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,12288,0.01310400002532535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,8192,0.009563555320103964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,10240,0.012061333490742577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,12288,0.012815111213260226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,7168,0.008047999607192146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,1536,0.006030222194062338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,10240,0.011696000066068439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,8192,0.010739555789364709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,6144,0.00903733323017756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,5120,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,7168,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,8192,0.010399111443095738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,512,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,7168,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,4096,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,5120,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,3584,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,5120,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,4096,0.008295999632941352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,3072,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,4096,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,3584,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,2560,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,3584,0.007368000017272101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,3072,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,3072,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,2560,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,2560,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,1536,0.0038515557017591265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,2048,0.006706666615274217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,2048,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,1024,0.0031759999692440033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,1536,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,1536,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,768,0.002834666727317704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,1024,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,1024,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,6144,0.009682666924264696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,512,0.0025235555238193935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,6144,0.009229333036475712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,768,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,256,0.0021688888470331826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,768,0.0053031109273433685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,512,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,128,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,512,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,256,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,64,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,32,0.0018595556418100994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,128,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,65536,0.021104888783560857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,16384,0.010969777901967367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,128,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,768,2048,0.004507555729813046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,51200,0.017332444588343304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,16384,0.011659555964999728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,65536,0.05272000034650167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,51200,0.04208533300293816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,65536,0.04986933204862806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,12288,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,51200,0.04006310966279771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,16384,0.01591999994383918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,16384,0.015449777245521545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,12288,0.013220444321632385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,8192,0.007991111112965478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,12288,0.012755555411179861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,10240,0.0120328888297081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,7168,0.007600000335110559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,10240,0.011696000066068439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,8192,0.010670222342014313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,6144,0.009360888765917884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,8192,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,7168,0.010294222169452244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,5120,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,7168,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,6144,0.00943288869327969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,4096,0.006903111106819577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,256,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,5120,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,3584,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,5120,0.008273777862389883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,4096,0.008040889269775814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,64,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,3072,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,4096,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,3584,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,2560,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,3584,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,3072,0.007490666376219855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,2560,0.007255111303594377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,3072,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,10240,0.008626666333940294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,1536,0.003830222205983268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,2560,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,1024,0.0031582222630580268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,2048,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,2048,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,1536,0.006568888823191325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,768,0.002847111059559716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,1536,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,1024,0.006228444476922353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,512,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,768,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,1024,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,256,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,768,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,512,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,128,0.0019128889673286015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,512,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,64,0.0018568889548381169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,6144,0.009027555584907532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,128,0.005547555370463266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,32,0.001855111163523462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,256,0.005057777795526716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,65536,0.018191999859280057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,128,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,51200,0.01573333309756385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,64,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,32,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,65536,0.046909332275390625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,32,0.004871110949251386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,16384,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,65536,0.049031111929151744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,512,2048,0.004237333519591225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,51200,0.03760000069936117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,16384,0.015784000356992085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,10240,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,12288,0.013220444321632385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,16384,0.01532977819442749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,8192,0.007420444654093848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,10240,0.012081777883900536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,7168,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,10240,0.011502222054534487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,8192,0.010675555302037133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,6144,0.007188444336255391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,8192,0.010068444742096795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,7168,0.00999466660949919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,7168,0.009528000321653154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,5120,0.006567111031876669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,6144,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,4096,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,6144,0.008972444468074376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,5120,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,3584,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,5120,0.008351999852392409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,4096,0.008052444292439355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,3072,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,4096,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,3584,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,2560,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,3584,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,3072,0.007522666619883643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,2048,0.0041804446114434134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,51200,0.03804088963402642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,2560,0.007254222200976477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,1536,0.003547555456558863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,2560,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,12288,0.012844444149070315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,1024,0.003035555490189128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,2048,0.0069004446268081665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,768,0.0027635554886526535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,2048,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,1536,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,512,0.002493333278430833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,768,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,1024,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,256,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,512,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,512,0.005032000028424793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,128,0.0018648888087934917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,256,0.005069333232111401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,64,0.0018159999615616268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,128,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,32,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,128,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,256,12288,0.009002666506502364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,65536,0.023429332507981196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,64,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,3072,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,51200,0.020384889509942796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,65536,0.046224001381132335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,32,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,16384,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,65536,0.04721777637799581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,51200,0.03729866610633002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,12288,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,51200,0.03729866610633002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,16384,0.01591999994383918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,1024,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,10240,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,16384,0.01517955627706316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,12288,0.013423999978436364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,12288,0.012836444709036084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,10240,0.011910222470760345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,7168,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,8192,0.010703999963071613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,10240,0.011432888607184092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,6144,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,8192,0.010275555153687796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,7168,0.010053333308961656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,5120,0.00655377780397733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,6144,0.009462222456932068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,4096,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,5120,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,5120,0.008283555507659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,3584,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,4096,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,4096,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,3072,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,3584,0.007894222107198503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,3584,0.007264888948864407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,1536,0.006555555595291986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,2560,0.0048071112897661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,3072,0.007000000112586551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,3072,0.007615110940403408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,2048,0.0041831110914548235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,2560,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,768,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,2560,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,2048,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,2048,0.0064035554726918536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,1024,0.0028320000403457214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,7168,0.009848888549539778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,1536,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,1536,0.0058782220714622075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,768,0.0026897777699761917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,6144,0.008805333740181392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,512,0.00254488881263468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,1024,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,768,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,256,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,768,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,512,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,128,0.0019084444890419643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,512,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,256,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,64,0.0017964444640609955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,256,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,128,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,32,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,128,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,65536,0.010430222584141625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,51200,0.009119110802809397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,8192,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,16384,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,64,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,12288,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,32,0.004613333278232151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,128,1536,0.00349866681628757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,10240,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,8192,0.0053635554181204895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,65536,0.046521776252322726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,7168,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,16384,0.01519555515713162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,6144,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,12288,0.012421333127551608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,5120,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,8192,0.010033778018421596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,1024,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,4096,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,7168,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,3584,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,3072,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,6144,0.008767111433876885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,5120,0.00810222244924969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,2560,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,2048,0.004544888933499654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,4096,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,3584,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,1536,0.003539555602603488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,1024,0.0028773333049482773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,3072,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,768,0.002521777732504739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,2560,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,512,0.0023635555472638872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,1536,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,256,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,1024,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,128,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,768,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,64,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,64,32,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,512,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,65536,0.011388444238238864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,51200,0.0373617774910397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,256,0.005065777649482091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,51200,0.009347555538018545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,128,0.005098666581842634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,64,0.00464088883664873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,12288,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,32,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,10240,0.011143111520343356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,10240,0.005623111294375525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,8192,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,65536,0.04617599977387322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,7168,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,51200,0.03730488816897074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,16384,0.01520088811715444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,6144,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,5120,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,10240,0.011196444431940714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,4096,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,8192,0.010097777677906884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,7168,0.009415999882751042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,6144,0.008792888787057664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,3072,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,5120,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,2560,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,2048,0.006224000205596288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,4096,0.0074542222751511475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,2048,0.004552888787455029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,3584,0.007244444555706448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,1536,0.003481777591837777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,3072,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,1024,0.0028239999794297745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,2048,0.006292444550328785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,768,0.00254133323000537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,1536,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,512,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,16384,0.005985777825117111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,256,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,1024,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,128,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,768,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,64,0.001818666648533609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,512,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,32,0.001818666648533609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,12288,0.012411555482281579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,256,0.004966222163703707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,128,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,81,32,3584,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,64,0.0047173334492577445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,32,0.004563555535342959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,16384,0.3662782245212131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,16384,0.2009742259979248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,51200,0.6044453514946831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,12288,0.27558665805392796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,16384,0.14034044742584229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,12288,0.1486862235599094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,51200,0.39438843727111816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,2560,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,10240,0.22049778037601045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,12288,0.10039822260538737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,10240,0.12669866614871556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,51200,1.1205964618259006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,10240,0.08462933037016128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,8192,0.17375644048055014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,16384,0.28772444195217556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,7168,0.15380444791581896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,12288,0.21823378403981528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,8192,0.07505955298741658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,8192,0.10727910863028632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,6144,0.13026666641235352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,7168,0.09188533491558498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,10240,0.1808373398251004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,7168,0.06648888852861193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,5120,0.11358400185902913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,6144,0.0786000026596917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,8192,0.1471911138958401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,6144,0.052298665046691895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,5120,0.0660746693611145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,4096,0.09029510948393081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,5120,0.0466284453868866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,7168,0.1307671070098877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,4096,0.0554346674018436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,3584,0.07999822166230944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,4096,0.03891822364595201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,6144,0.11448000537024604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,5120,0.09486844142278035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,3584,0.048730668094423085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,3584,0.03528711199760437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,4096,0.07932266924116346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,2560,0.05866489145490858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,3072,0.042380445533328585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,3072,0.03087110983000861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,3584,0.07061866919199626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,2048,0.04778755704561869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,2560,0.03695733348528544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,3072,0.06461955441368951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,2560,0.026757333013746474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,1536,0.03664977682961358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,2048,0.031919111808141075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,2048,0.02461066676510705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,1024,0.027425777581002977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,2560,0.05592444207933214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,1536,0.03959111041492886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,1536,0.018434666925006442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,768,0.020590222544140287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,2048,0.048844443427191846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,1024,0.015056888262430826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,1536,0.04264355699221293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,1024,0.03014755580160353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,512,0.013776889277829064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,3072,0.0681848857137892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,768,0.013541333377361298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,768,0.023256889647907678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,1024,0.03671377897262573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,256,0.008076444268226624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,512,0.01714666684468587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,256,0.011038222246699862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,768,0.028913779391182795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,128,0.00655377780397733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,512,0.012191111014948951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,64,0.0042053332759274375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,512,0.024705777565638225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,256,0.020541333489947848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,32,0.005439110928111606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,51200,0.8777208858066134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,128,0.011370666325092316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,64,0.011749332977665795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,128,0.015955555770132277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,32,0.01551466683546702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,65536,0.6403875350952148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,256,0.012429333395428128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,65536,1.0880230797661674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,51200,0.8514977561102973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,128,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,16384,0.15939822461869982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,16384,0.27686044904920787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,65536,0.401256004969279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,51200,0.4753031200832791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,12288,0.20844177405039468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,12288,0.11786844995286729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,16384,0.11173866854773627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,10240,0.17434133423699272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,51200,0.31375734011332196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,12288,0.08124622371461657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,10240,0.10019199715720283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,10240,0.06823111242718168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,8192,0.1410533322228326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,12288,0.17709599600897896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,16384,0.2305448849995931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,7168,0.12503466341230604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,8192,0.05765599674648709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,10240,0.14720711443159315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,7168,0.07210933499866061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,6144,0.10620444350772434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,7168,0.05319911241531372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,8192,0.1196791066063775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,8192,0.08551910850736831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,5120,0.09072089195251465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,7168,0.1061555544535319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,6144,0.04433777928352356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,4096,0.07358577516343859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,5120,0.053063111172782056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,6144,0.0934257772233751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,5120,0.038147555457221136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,3584,0.06527111265394422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,4096,0.04430400000678169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,5120,0.08024266693327162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,3584,0.03954933418167962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,3584,0.029046220911873713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,3072,0.057051552666558154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,3584,0.06189777453740438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,3072,0.03457955519358317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,6144,0.0631093316608005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,2560,0.04775377776887682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,3072,0.025432889660199482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,3072,0.05521688858668009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,2048,0.03905511233541701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,2560,0.022535110513369244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,2560,0.03035911255412632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,51200,0.715093347761366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,1536,0.03083555565940009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,2560,0.048972444401847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,4096,0.03225866622394986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,2048,0.026532444689008925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,1024,0.022583999567561682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,2048,0.020356444848908316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,1536,0.01726222203837501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,4096,0.06901866859859891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,2048,0.04278666774431864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,1536,0.03285955720477634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,768,0.01610133383009169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,1024,0.013373333546850415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,1536,0.03641511003176371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,768,0.012578667038016848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,768,0.01891644464598762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,1024,0.031713777118259005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,256,0.0074471111098925276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,512,0.011042666931947073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,768,0.024857777688238356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,256,0.0107360002067354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,512,0.021749332547187805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,256,0.009690666364298927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,128,0.004522666749027041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,128,0.009933333430025313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,64,0.0038559999730851916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,128,0.010017777482668558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,32,0.00387999994887246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,65536,0.9542559517754449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,64,0.009992000129487779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,128,0.014154665999942355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,1024,0.023754666248957317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,512,0.01128533316983117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,32,0.013280889226330651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,65536,0.3238995605044895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,512,0.013991110854678683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,51200,0.24920177459716797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,65536,0.22365066740247938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,16384,0.08214400211970012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,51200,0.17156355910831025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,65536,0.1488702164755927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,12288,0.06503378020392524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,16384,0.06016977628072103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,16384,0.04376088910632663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,51200,0.11212889353434245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,256,0.01828799976242913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,10240,0.05280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,12288,0.04559822214974297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,12288,0.03329688972897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,10240,0.039376888010236956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,16384,0.060136000315348305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,8192,0.044806222120920815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,12288,0.04658755660057068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,10240,0.04109777675734626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,7168,0.03851111067665948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,8192,0.03465955456097921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,8192,0.023911999331580266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,8192,0.03602044449912177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,7168,0.029559999704360962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,6144,0.033442666133244835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,7168,0.023019555542204116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,7168,0.03216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,6144,0.026403556267420452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,5120,0.028764443265067205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,6144,0.029829333225886028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,5120,0.022800000177489385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,10240,0.028903110159768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,4096,0.023825777901543513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,5120,0.017151110702090792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,5120,0.026368889543745253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,3584,0.02092622220516205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,4096,0.015365334020720588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,4096,0.01925155520439148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,51200,0.15980088710784912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,4096,0.022779555784331426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,3584,0.017298666967286002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,3072,0.018437332577175565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,65536,0.20390400621626112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,2560,0.015470221638679504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,3584,0.021244444780879553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,6144,0.019374221563339233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,3072,0.015490666031837463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,2048,0.012423999607563019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,2560,0.013802666631009845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,3072,0.01920977731545766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,1536,0.010079999764760336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,2560,0.017903111047214933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,2048,0.01143555591503779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,1536,0.009744000103738572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,1024,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,1536,0.013844444519943662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,1536,0.01535555554760827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,3584,0.014523555835088095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,1024,0.011790222591824003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,768,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,1024,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,3072,0.013174222575293647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,512,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,768,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,768,0.009389333426952362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,2560,0.012044444680213928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,512,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,256,0.0031582222630580268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,512,0.008151111503442129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,2048,0.012422222230169507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,128,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,256,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,512,0.010779555473062726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,256,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,64,0.002495999965402815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,2048,0.01648533344268799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,128,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,32,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,256,0.009667555491129557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,128,0.006607111129495833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,64,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,32,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,65536,0.25137244330512154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,128,0.008285333712895712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,65536,0.1727093325720893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,1024,0.014072888427310519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,51200,0.1956311066945394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,65536,0.1306017769707574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,768,0.012755555411179861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,51200,0.10054400232103135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,12288,0.054710222615136035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,16384,0.038316445218192205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,16384,0.05162488751941257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,12288,0.03625066743956672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,10240,0.045019557078679405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,12288,0.02954933378431532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,65536,0.16552000575595432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,12288,0.04110222061475118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,51200,0.13328088654412165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,8192,0.03799555699030558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,10240,0.026296888788541157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,10240,0.03137244449721442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,16384,0.06992089086108737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,7168,0.03307644526163737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,8192,0.021225778592957392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,10240,0.03644355469279819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,51200,0.13253866301642522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,6144,0.028566221396128338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,16384,0.047127111090554126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,7168,0.019747555255889893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,8192,0.032123555739720665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,7168,0.02401599950260586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,5120,0.024504888388845656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,6144,0.02121066715982225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,7168,0.029048889875411987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,6144,0.018002667360835604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,4096,0.021273778544531927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,5120,0.018170666363504197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,6144,0.027092445227834914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,5120,0.015832000308566622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,3584,0.018208000395033095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,4096,0.01568888955646091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,5120,0.023731556203630235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,4096,0.01463733282354143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,3072,0.015568888849682279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,3584,0.013431111143694984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,4096,0.020799999435742695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,3584,0.014431110686726041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,2560,0.013048888908492194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,3072,0.013091555900043912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,3072,0.01258933295806249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,3584,0.019468444916937087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,2048,0.011175110936164856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,8192,0.02717422280046675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,2560,0.011730666789743634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,2560,0.011970666547616323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,3072,0.01809777816136678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,1536,0.009222221871217092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,2048,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,2048,0.011050666371981302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,1024,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,1536,0.00868977771864997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,1536,0.0123857781291008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,2048,0.015831111205948722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,768,0.006118222243256039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,1024,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,1536,0.014801777071423002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,1024,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,512,0.004841777599520153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,768,0.0073928890956772705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,768,0.008416000339719985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,1024,0.013078221844302284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,256,0.0035342222286595237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,512,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,512,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,768,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,128,0.0028373334142896864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,256,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,512,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,64,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,32,0.002587555597225825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,128,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,128,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,256,0.009642666412724389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,64,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,128,0.008013333711359236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,65536,0.21377600563897026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,2560,0.016937777400016785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,51200,0.16658666398790148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,65536,0.12466400199466282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,51200,0.12639821900261775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,16384,0.05925866630342272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,51200,0.09516177574793498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,16384,0.04505866765975952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,12288,0.04585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,16384,0.03512800070974562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,256,0.006261333409282897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,16384,0.049215111467573375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,12288,0.028373334142896865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,65536,0.15551555156707764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,51200,0.12394932905832927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,12288,0.03464711043569777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,8192,0.03237511052025689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,32,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,12288,0.039444443252351545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,10240,0.024159111910396155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,10240,0.030151999659008447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,65536,0.16447910997602674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,7168,0.028093334701326158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,8192,0.026068444053332012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,8192,0.022042666872342426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,6144,0.023936000135209825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,10240,0.035042666726642184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,7168,0.019759999381171334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,5120,0.02087200019094679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,7168,0.022674666510687932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,6144,0.017403556240929496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,6144,0.02016799979739719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,7168,0.027263111538357202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,4096,0.01776977711253696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,5120,0.015616888801256815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,5120,0.017161778277821012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,6144,0.025850666893853083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,3584,0.015121777852376303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,4096,0.01401422255569034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,5120,0.02273955610063341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,4096,0.015231110983424716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,3072,0.013785777820481194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,10240,0.03768800033463372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,3584,0.01309511148267322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,4096,0.02033066584004296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,2560,0.011952000359694162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,3072,0.012738666600651212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,3072,0.012198222180207571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,3584,0.019278221660190158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,2048,0.009713778065310584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,2560,0.011704000333944956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,2560,0.01068800025516086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,3072,0.017796445224020217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,1536,0.008268444074524773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,2560,0.016354666815863717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,2048,0.00923644420173433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,8192,0.030925333499908447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,1024,0.007108444141017065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,1536,0.011192888849311404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,1536,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,2048,0.015486222174432544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,768,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,1024,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,1024,0.008829333715968663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,1536,0.014160000615649752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,512,0.004754666652944353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,768,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,768,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,1024,0.012116444607575735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,3584,0.013928888572586907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,256,0.0033004445334275565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,512,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,512,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,128,0.002838222309947014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,256,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,256,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,512,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,64,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,256,0.01000088867213991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,128,0.006317333214812809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,2048,0.010567111273606619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,64,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,128,0.007903999752468532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,65536,0.16986400551266143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,32,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,51200,0.13297866450415716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,65536,0.12060177326202393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,65536,0.07113777928882174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,768,0.011359111302428775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,16384,0.046738667620552905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,51200,0.05704622136221992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,128,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,32,0.002521777732504739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,65536,0.13966755072275797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,16384,0.03456266721089681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,16384,0.023056889573733013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,51200,0.11156888802846272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,12288,0.018555555078718398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,16384,0.04498844345410665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,12288,0.03696622120009528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,10240,0.01629866659641266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,8192,0.025780444343884785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,10240,0.0329039990901947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,51200,0.09288177887598674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,8192,0.020290666156344943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,7168,0.02214755614598592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,8192,0.014465777410401238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,8192,0.028174221515655518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,6144,0.01939733326435089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,7168,0.01351111133893331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,7168,0.017626666360431247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,12288,0.036159111393822566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,5120,0.016513778103722464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,7168,0.026847999956872728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,6144,0.01607466737429301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,12288,0.026770666241645813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,4096,0.014138667119873894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,5120,0.014090667168299357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,6144,0.023959999283154804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,5120,0.011340444286664328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,3584,0.01240799989965227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,10240,0.02340000040001339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,5120,0.021670222282409668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,4096,0.012744888663291931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,3072,0.011006222003036074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,3584,0.009665778113736046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,4096,0.019800888167487252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,3584,0.01832799944612715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,3072,0.010968888799349466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,2560,0.009282666775915358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,3072,0.017464000317785475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,2560,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,6144,0.012383999923865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,2048,0.008309333688682979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,10240,0.030344890223609075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,2560,0.008280889027648503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,1536,0.006791111081838608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,2560,0.01612711118327247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,2048,0.007632888853549957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,2048,0.009144888983832465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,1024,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,1536,0.00720088928937912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,1536,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,4096,0.010441777606805166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,2048,0.015237333046065437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,768,0.004506666627195147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,1024,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,3584,0.011766222616036734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,1536,0.013658666776286231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,1024,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,512,0.0034248888906505373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,768,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,768,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,3072,0.008783111141787635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,1024,0.011237333218256632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,512,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,768,0.010899555351999072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,512,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,128,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,256,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,512,0.01033688916100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,256,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,64,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,32,0.0024764444679021835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,128,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,128,0.005372444374693765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,256,0.009338666995366415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,128,0.007583111524581909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,64,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,65536,0.14952266216278076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,32,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,51200,0.1193457841873169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,65536,0.11827199988894993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,16384,0.04547911220126682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,65536,0.0681555536058214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,51200,0.09124622080061172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,51200,0.055352889829211764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,12288,0.03346844514211019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,256,0.0027751111321979095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,16384,0.02167555524243249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,16384,0.03345600101682875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,10240,0.027785778045654297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,65536,0.13752355840471056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,12288,0.01794488893614875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,16384,0.04457599918047587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,51200,0.11055733097924127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,12288,0.02590933276547326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,8192,0.02334044377009074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,10240,0.015810666812790763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,10240,0.022384888596004907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,12288,0.03657777773009406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,8192,0.01407999959256914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,8192,0.019494222270117868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,6144,0.0181057784292433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,10240,0.03191377719243368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,7168,0.013099555340078143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,7168,0.017267554998397827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,8192,0.028355555401908025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,5120,0.015434667468070984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,6144,0.012144889268610211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,7168,0.025451555848121643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,6144,0.01568177839120229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,4096,0.013220444321632385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,5120,0.01385155568520228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,5120,0.011095999843544431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,3584,0.011764444410800934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,6144,0.023312888211674158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,4096,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,5120,0.021332444416152105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,4096,0.012588444683286877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,3072,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,3584,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,2560,0.009183111290136972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,3584,0.011686222420798408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,4096,0.019311111834314134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,3072,0.008354666332403818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,2048,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,3584,0.01829600003030565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,2560,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,7168,0.02055644492308299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,2560,0.009825777676370408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,3072,0.017250667015711468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,2048,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,2048,0.008588444855478074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,2560,0.016167110866970487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,1024,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,1536,0.007000889215204451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,1536,0.008824889030721452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,768,0.0038657776183552216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,1024,0.007744889292452071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,1536,0.012375999655988483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,1024,0.011066666907734342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,768,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,512,0.0032257777121331957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,768,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,512,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,768,0.01073333372672399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,256,0.0026488889836602737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,512,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,512,0.009780444204807281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,3072,0.010687111152542962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,256,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,128,0.002492444382773505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,256,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,1536,0.006568888823191325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,64,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,128,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,32,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,256,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,128,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,65536,0.12508711549970838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,64,0.005624889085690181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,128,0.007252444823582967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,32,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,1024,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,51200,0.09956977764765422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,65536,0.11321777767605251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,65536,0.06567555665969849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,16384,0.039001779423819646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,51200,0.08641422457165188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,16384,0.021294222937689886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,12288,0.03191555539766947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,51200,0.11019644472334121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,65536,0.13724799950917563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,16384,0.04411377840571933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,10240,0.027779555983013574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,12288,0.02462399999300639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,2048,0.014926221635606555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,10240,0.02132355504565769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,12288,0.035855998595555626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,8192,0.021031111478805542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,10240,0.01536711057027181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,10240,0.03140177660518222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,7168,0.018607111440764535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,8192,0.01868711080816057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,51200,0.05316088928116692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,7168,0.012663111090660095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,8192,0.027687110834651526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,16384,0.03214844399028354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,7168,0.016770665844281513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,6144,0.01815822223822276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,5120,0.015040889382362366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,7168,0.024826667375034753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,6144,0.011674666570292579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,6144,0.015433778365453085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,4096,0.012797333300113678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,5120,0.01351999988158544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,12288,0.0173680004146364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,6144,0.02330133318901062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,3584,0.01089244418674045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,4096,0.009195555415418413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,4096,0.012096000214417776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,5120,0.020921778347757127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,3072,0.01068800025516086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,3584,0.011362666885058085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,3584,0.008629333641793992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,4096,0.018923555811246235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,2560,0.00869688888390859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,8192,0.01351466692156262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,3072,0.010631110933091907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,3072,0.008368888662921058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,2048,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,2560,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,2560,0.009688888986905416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,1536,0.006575999988449945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,2048,0.008014221986134848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,2048,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,2560,0.015417777829700045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,1024,0.004855111241340637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,1536,0.006903111106819577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,2048,0.014127999544143677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,1536,0.008294222255547842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,768,0.003751111113362842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,5120,0.010634666515721215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,1024,0.006503110958470239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,1024,0.007159111400445302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,512,0.003165333428316646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,768,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,768,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,1024,0.011408888631396823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,256,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,512,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,512,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,768,0.01088444474670622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,128,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,256,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,256,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,512,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,64,0.0022008888837363985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,3072,0.016892444756295946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,128,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,32,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,128,0.005879111174080107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,256,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,65536,0.11721600426567926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,64,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,128,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,32,0.005981333139869902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,51200,0.09073955482906765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,65536,0.10792266660266453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,65536,0.06336710850397746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,16384,0.03295288814438714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,1536,0.0123831108212471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,51200,0.08367733160654704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,51200,0.051144000556733876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,12288,0.02736088964674208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,16384,0.020271999968422782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,65536,0.13709244463178846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,16384,0.030918220678965252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,51200,0.10978044403923883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,10240,0.023727999793158636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,12288,0.01676977839734819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,16384,0.04387911160786947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,3584,0.017836444907718234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,8192,0.0211173329088423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,10240,0.014939554863505893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,12288,0.03556088937653436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,10240,0.020777778493033517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,7168,0.01908622185389201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,8192,0.01817244456874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,8192,0.013179555535316467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,10240,0.030885332160525854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,7168,0.016340444485346477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,7168,0.012429333395428128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,8192,0.02701777716477712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,5120,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,6144,0.015064888530307345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,7168,0.024832000335057575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,5120,0.013239111337396832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,5120,0.010037333601050908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,4096,0.009988444546858469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,6144,0.023063111636373732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,4096,0.009015999734401703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,5120,0.0207306659883923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,12288,0.023638221952650283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,3584,0.009176889227496253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,4096,0.011917333636018964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,3072,0.010608000059922537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,3584,0.008232889076073965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,4096,0.01900888813866509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,3584,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,2560,0.009160888691743216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,6144,0.01350488927629259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,3072,0.009927111367384592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,3072,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,2048,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,2560,0.008910222185982598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,3072,0.01694222291310628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,6144,0.011109333071443768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,1536,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,2048,0.00794666674402025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,2048,0.007191110816266801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,2560,0.01478488908873664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,1024,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,1536,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,2048,0.013207111093733044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,1536,0.007671111159854465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,768,0.0038986665507157645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,1536,0.012068444656001197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,1024,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,768,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,768,0.006224000205596288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,1024,0.011400000088744693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,512,0.0035217776894569397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,256,0.0028355556229750314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,512,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,768,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,512,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,128,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,256,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,512,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,256,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,64,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,128,0.0058711109062035876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,256,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,128,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,32,0.002481777841846148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,65536,0.08612889051437378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,128,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,32,0.0058444444504049085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,1024,0.006360888895061281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,65536,0.06531200144026014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,51200,0.07021688752704196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,65536,0.061214221848381885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,16384,0.028500444359249536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,51200,0.05195111036300659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,3584,0.01733955575360192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,65536,0.13675467173258463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,16384,0.019925332731670804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,12288,0.023000000251664057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,16384,0.02012977831893497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,12288,0.01621955633163452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,16384,0.042974223693211876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,12288,0.016176000237464905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,10240,0.019873778025309246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,10240,0.014489778214030795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,12288,0.03460444344414605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,64,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,8192,0.015098666151364645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,10240,0.014546665880415173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,7168,0.013425778183672162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,8192,0.012859555582205454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,10240,0.03029600116941664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,8192,0.013105777402718862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,6144,0.011719110939237805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,7168,0.011772444678677453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,7168,0.011672000090281168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,51200,0.048911998669306435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,8192,0.02681511143843333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,5120,0.010615111225181155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,6144,0.010515555739402771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,6144,0.011046222514576383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,7168,0.02435644467671712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,2560,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,4096,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,5120,0.01000355515215132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,5120,0.009482666850090027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,6144,0.022690667046440974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,3584,0.008092444803979661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,4096,0.008348444269763099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,4096,0.00907022257645925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,3072,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,5120,0.020238222347365487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,3584,0.0084915558497111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,3584,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,4096,0.0184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,2560,0.006779555645253923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,3072,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,3584,0.017264000243610807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,3072,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,2048,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,2560,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,2560,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,1536,0.004685333205593957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,3072,0.015787555111779105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,2048,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,2560,0.01477599971824222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,1024,0.0035111109415690103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,2048,0.013202666408485837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,1536,0.006826666494210561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,51200,0.1094542211956448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,768,0.0032133333798911837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,1024,0.006346666564544042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,1024,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,768,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,512,0.0028275555620590844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,768,0.005662222289376789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,1024,0.011004444625642566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,512,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,768,0.010064889159467485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,256,0.0024480000138282776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,512,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,256,0.0057911111248864075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,512,0.01034755590889189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,256,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,128,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,2048,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,64,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,128,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,256,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,1536,0.006690666493442323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,65536,0.07664444711473253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,64,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,1536,0.012048000262843238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,128,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,32,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,51200,0.061998221609327525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,65536,0.059918224811553955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,65536,0.0644213358561198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,16384,0.02514666650030348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,51200,0.0500257776843177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,51200,0.04780088861783346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,12288,0.019353777170181274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,16384,0.019357333580652874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,16384,0.019433778193261888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,51200,0.10913155476252238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,10240,0.016553777787420485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,12288,0.015559110376569959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,12288,0.016014221641752455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,16384,0.04248977700869242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,8192,0.015842666228612263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,10240,0.014172444740931192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,128,0.004987555659479565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,12288,0.03432800041304694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,10240,0.013912889692518445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,7168,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,8192,0.012601777911186218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,10240,0.030060443613264296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,6144,0.01276533305644989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,7168,0.011416888899273343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,7168,0.011045333411958484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,8192,0.026370666093296472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,5120,0.009917333722114563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,7168,0.02438755498992072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,6144,0.011010666688283285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,5120,0.00980177770058314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,5120,0.009310222334331935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,65536,0.13667110602060953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,4096,0.008968888885445064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,6144,0.022170666191313002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,3584,0.007872000336647034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,4096,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,5120,0.020357333951526217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,4096,0.008427555362383524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,3072,0.007255111303594377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,32,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,3584,0.007707555260923173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,4096,0.018311111463440787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,8192,0.012732444538010491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,3072,0.007718222008811102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,2560,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,3584,0.016992888516849942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,3072,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,2048,0.005311111195219888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,3072,0.015450666348139444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,2560,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,2560,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,6144,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,1536,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,2048,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,2560,0.014030221435758801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,1024,0.0034844444857703317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,1536,0.006663110935025745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,1536,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,2048,0.013539555172125498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,768,0.0031617778456873367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,1536,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,1024,0.0058746664888328975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,3584,0.00814400033818351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,768,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,768,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,1024,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,512,0.00281866660548581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,256,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,512,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,768,0.010056888891590966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,512,0.005379555539952383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,128,0.002130666747689247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,256,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,256,0.005792888916201062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,512,0.01035288886891471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,64,0.002143111079931259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,2048,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,128,0.005766222046481238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,256,0.0090506664580769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,32,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,128,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,65536,0.06663288672765096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,64,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,128,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,32,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,51200,0.053371555275387235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,65536,0.06179022126727634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,1024,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,65536,0.05881777736875746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,16384,0.02071199980047014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,51200,0.04665066798528036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,51200,0.04805155595143636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,12288,0.01681511104106903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,16384,0.01919555498494042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,16384,0.018745778335465323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,10240,0.014525334040323893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,12288,0.015757333901193406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,51200,0.10897066858079697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,12288,0.015156444576051501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,16384,0.04225511021084256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,8192,0.012764444781674279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,10240,0.01404888927936554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,12288,0.03412088751792908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,7168,0.012288000020715924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,8192,0.011687111523416309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,8192,0.012362666428089142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,6144,0.011071110765139261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,7168,0.010694222317801582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,7168,0.01106844428512785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,6144,0.010407999985747868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,7168,0.024511110451486375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,6144,0.009726222190592024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,65536,0.13642222351498073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,6144,0.02205955485502879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,4096,0.008630221916569604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,5120,0.009677333136399588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,5120,0.008959111240175035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,5120,0.019820445113711886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,3584,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,4096,0.008296888735559251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,4096,0.00865955568022198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,10240,0.013614222407341003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,3072,0.0068808889223469635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,3584,0.008094222181373173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,4096,0.017864889568752713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,3584,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,10240,0.02990666694111294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,2560,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,3072,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,3072,0.007337777978844113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,3584,0.01607111096382141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,2048,0.005375999957323074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,8192,0.026053332620196875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,2560,0.007344000041484833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,2560,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,1536,0.004512000001139111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,5120,0.009646221995353699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,3072,0.014840000205569796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,1024,0.0034391110142072043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,2048,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,2560,0.014103111293580798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,1536,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,1536,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,2048,0.013432888521088494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,768,0.0029351111087534162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,1024,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,512,0.0027431110954946945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,1536,0.012098666694429187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,768,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,1024,0.010682666467295753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,768,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,256,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,512,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,768,0.010099555883142684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,128,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,256,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,256,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,512,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,64,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,128,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,256,0.009125333693292405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,128,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,65536,0.05827022261089749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,2048,0.007176000210973952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,64,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,128,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,32,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,51200,0.04667822188801236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,65536,0.05779111385345459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,65536,0.05952711237801445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,16384,0.019111999207072787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,51200,0.04620000057750278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,51200,0.04549244377348158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,65536,0.13612711429595947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,16384,0.018000000052981906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,51200,0.10912977986865574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,16384,0.04166577921973334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,12288,0.0148026661740409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,10240,0.014105778601434497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,12288,0.015757333901193406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,32,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,8192,0.012378666963842181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,10240,0.013391999734772576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,10240,0.01315199997689989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,8192,0.012244444754388599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,10240,0.029819554752773706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,7168,0.011769777370823754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,8192,0.01108088923825158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,8192,0.025722665919197932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,7168,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,6144,0.010327111515733931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,7168,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,12288,0.01628800067636702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,5120,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,6144,0.00980355590581894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,6144,0.009962666365835402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,5120,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,6144,0.021703110800849065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,4096,0.008072000410821702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,5120,0.00889511075284746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,12288,0.03396711084577773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,4096,0.008275555239783393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,5120,0.019832000136375427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,3584,0.007280888656775157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,4096,0.007870222131411234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,3072,0.006894222150246303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,3584,0.008174222376611497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,4096,0.01753777762254079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,3584,0.007596444752481248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,2560,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,3072,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,3072,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,3584,0.01591466698381636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,2048,0.004868444469239977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,2560,0.007328888608349695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,2560,0.006697777658700943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,512,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,1536,0.0041946665280395085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,7168,0.02370044423474206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,2048,0.006312888943486744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,2048,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,2560,0.01407999959256914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,1024,0.00319377767542998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,1536,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,1536,0.0060159998635451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,2048,0.012840000291665396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,768,0.0029324444217814338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,1024,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,1536,0.011743110915025076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,512,0.0024862223201327855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,768,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,768,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,768,0.010210666391584609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,512,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,512,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,256,0.002181333386235767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,256,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,128,0.002191999927163124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,512,0.009712888962692684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,256,0.005098666581842634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,16384,0.018910222583346896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,256,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,64,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,128,0.005730666634109285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,128,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,32,0.001963555605875121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,65536,0.04829688866933187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,128,0.007346666521496243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,64,0.00498844434817632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,1024,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,51200,0.038731555143992104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,65536,0.05608355336719089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,1024,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,65536,0.056200888421800405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,16384,0.016362667083740234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,51200,0.04451733496454027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,51200,0.0450764430893792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,16384,0.01774311065673828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,16384,0.017447110679414537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,12288,0.013415111435784234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,51200,0.10929244756698608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,3072,0.015396444333924187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,10240,0.012791111237472959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,16384,0.041249778535630964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,12288,0.014736889137162102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,12288,0.014688000082969666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,8192,0.011393778026103973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,10240,0.012368000215954252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,10240,0.01310400002532535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,12288,0.03376444511943393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,7168,0.010703999963071613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,32,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,8192,0.011648000114493899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,8192,0.011046222514576383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,6144,0.00980088859796524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,10240,0.029501332177056208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,7168,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,7168,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,5120,0.008723555339707268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,8192,0.02613777750068241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,6144,0.009646221995353699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,6144,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,7168,0.023676445086797077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,4096,0.009394666386975182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,5120,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,5120,0.008629333641793992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,6144,0.021593777669800654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,3584,0.00889866633547677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,4096,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,4096,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,5120,0.019317333896954853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,3584,0.007370666497283512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,4096,0.017218665944205392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,2560,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,3584,0.015864888827006023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,3072,0.007054222126801808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,2560,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,2048,0.004552888787455029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,3072,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,2560,0.013767110804716745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,65536,0.136080887582567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,2048,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,1536,0.0038906666967603895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,2048,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,2048,0.012995555169052549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,1024,0.0032044444233179092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,1536,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,1536,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,3072,0.007717333734035492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,768,0.0029395555870400537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,1024,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,1536,0.011741333537631564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,3584,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,1024,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,512,0.002691555561290847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,768,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,3072,0.007675555845101674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,768,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,256,0.002463999928699599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,512,0.005658666706747479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,512,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,2560,0.0068417779273457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,128,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,768,0.010407110883129967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,512,0.009682666924264696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,256,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,64,0.002201777779393726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,32,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,128,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,128,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,256,0.008989333278603023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,65536,0.03679377834002177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,64,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,128,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,32,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,51200,0.030117332935333252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,65536,0.055108444558249585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,65536,0.054048889213138156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,51200,0.04467733369933235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,51200,0.04309155543645223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,65536,0.13577244016859266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,16384,0.01721155477894677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,12288,0.013419555293189155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,16384,0.017163554827372234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,51200,0.1081635554631551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,16384,0.04142755601141188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,12288,0.014160000615649752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,10240,0.011944000091817645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,256,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,10240,0.012059555285506778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,8192,0.011169777976142036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,10240,0.012696888711717395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,10240,0.029765334394243028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,16384,0.01590755581855774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,7168,0.0101724440852801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,8192,0.010976889067225985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,8192,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,8192,0.02608266638384925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,6144,0.00943288869327969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,7168,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,7168,0.01001422190003925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,1024,0.010695111420419482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,5120,0.008912888665994009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,7168,0.023020444644822016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,6144,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,6144,0.009353777600659264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,12288,0.01441333360142178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,4096,0.0076248885856734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,5120,0.008762666748629676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,5120,0.008583111067612966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,3584,0.006872889068391588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,4096,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,12288,0.033542222446865506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,4096,0.007678222325113084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,5120,0.018564444449212816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,3072,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,3584,0.007956444389290279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,4096,0.01722044414944119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,3584,0.007345777418878343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,2560,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,3072,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,3584,0.016139555308553908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,2048,0.0042275554604000514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,3072,0.014803555276658801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,2560,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,2560,0.006691555596060223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,1536,0.004530666602982415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,2560,0.014144889182514615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,2048,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,2048,0.006411555740568373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,1024,0.0031795555518733132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,1536,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,2048,0.012772444221708508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,768,0.0028853333658642242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,1024,0.006303999986913469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,1024,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,1536,0.012346666720178394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,512,0.002516444358560774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,768,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,768,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,512,0.005649777750174205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,768,0.010049777726332346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,256,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,3072,0.007024889190991719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,256,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,512,0.009716444545321995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,128,0.0018942221585247253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,256,0.005173333403136996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,128,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,256,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,64,0.0019208888212839763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,128,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,32,0.001793777777089013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,1536,0.006572444405820634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,65536,0.026375111606385972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,128,0.007465778125656976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,64,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,6144,0.020571556356218126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,32,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,51200,0.02459200057718489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,65536,0.05163377854559156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,65536,0.05440444416469998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,16384,0.011238222320874533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,51200,0.04399822155634562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,51200,0.04127466678619385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,65536,0.13533955150180393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,12288,0.009529777699046666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,512,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,16384,0.01685599982738495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,16384,0.016199111938476562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,10240,0.010225777824719747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,51200,0.10766044590208267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,12288,0.013433777623706393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,16384,0.041466666592492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,8192,0.011439999772442712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,12288,0.013432888521088494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,10240,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,7168,0.011406222151385413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,12288,0.03323644399642944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,8192,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,8192,0.01034222212102678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,7168,0.009866666462686326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,7168,0.010072888599501716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,6144,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,7168,0.022664889693260193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,1024,0.011352000137170156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,6144,0.009325332939624786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,5120,0.008645333349704742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,6144,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,6144,0.020632889535692003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,5120,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,5120,0.008301333420806462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,5120,0.018921777606010437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,10240,0.012063110868136087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,3584,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,4096,0.008283555507659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,10240,0.029007110330793593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,4096,0.01649777756796943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,3072,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,3584,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,8192,0.025031111306614343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,3584,0.007678222325113084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,2560,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,3072,0.007589333587222629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,3584,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,3072,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,2048,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,2560,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,2560,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,3072,0.014763555592960782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,1536,0.004507555729813046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,4096,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,2048,0.0063493334584765965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,2560,0.014116444521480136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,1024,0.0042160000238153665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,1536,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,2048,0.013051555388503604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,4096,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,768,0.0034773333205117118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,1024,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,1024,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,1536,0.011696889168686338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,512,0.0028764444092909494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,768,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,1024,0.01070755554570092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,256,0.002453333387772242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,512,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,512,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,768,0.01036977767944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,128,0.0023271110322740343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,256,0.004991111242108875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,256,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,64,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,512,0.010030222435792288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,32,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,128,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,128,0.005528888768619961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,2048,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,65536,0.022096888886557683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,256,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,51200,0.02279377811484867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,128,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,64,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,32,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,65536,0.05422666668891907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,1536,0.005983110931184556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,16384,0.010775999890433418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,65536,0.05039733317163256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,51200,0.043453332450654775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,12288,0.009653333160612319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,51200,0.040639999839994646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,16384,0.01588711142539978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,10240,0.008712000317043727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,16384,0.0155102229780621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,65536,0.13606044981214735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,12288,0.013167999684810638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,51200,0.10764088895585801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,12288,0.013076444466908773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,8192,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,16384,0.04048533240954081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,10240,0.012058667010731168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,7168,0.009178666604889764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,10240,0.011726222104496427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,12288,0.03254222207599216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,8192,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,6144,0.008531555533409119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,8192,0.01036800030204985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,7168,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,10240,0.028470221492979262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,5120,0.00775733341773351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,7168,0.00979022185007731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,8192,0.024338665935728285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,6144,0.009016888837019602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,4096,0.006814222368929122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,7168,0.022575111852751836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,5120,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,6144,0.020288000504175823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,4096,0.008220444122950235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,4096,0.007675555845101674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,5120,0.018900444110234577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,4096,0.01646844380431705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,3584,0.00737955586777793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,3072,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,3584,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,3072,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,2560,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,3072,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,3072,0.014447111222479077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,2560,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,2048,0.004546666724814309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,2560,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,2560,0.013655111193656921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,5120,0.008837333156002892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,1536,0.0038497779104444715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,3584,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,2048,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,2048,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,1024,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,1536,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,2048,0.012418666647540199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,1536,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,768,0.0028071111689011254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,1024,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,1024,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,1536,0.011762667033407422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,512,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,768,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,768,0.0053324442770746015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,1024,0.011419555379284753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,256,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,512,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,512,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,768,0.010004444254769219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,128,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,6144,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,512,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,128,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,256,0.00965244405799442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,64,0.0018666666001081467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,768,32,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,65536,0.021095999413066443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,128,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,64,0.0052266667286554975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,32,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,51200,0.017450667089886136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,3584,0.015793777174419828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,65536,0.04963022139337328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,65536,0.05306755503018697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,16384,0.011006222003036074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,51200,0.038711110750834145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,51200,0.04001333316167196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,65536,0.13481956058078343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,12288,0.009390222529570261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,16384,0.015469332536061605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,16384,0.015789333317014907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,256,0.005025777965784073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,10240,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,12288,0.013419555293189155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,12288,0.012777778009573618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,128,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,16384,0.040236443281173706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,8192,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,10240,0.012031111452314587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,12288,0.0324577788511912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,7168,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,8192,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,10240,0.027790221903059218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,8192,0.02438844409253862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,7168,0.009710222482681274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,6144,0.0063528890411059065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,7168,0.010063110954231685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,7168,0.022297778063350256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,6144,0.009450666606426239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,5120,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,6144,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,256,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,5120,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,6144,0.020289777053727042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,4096,0.006340444501903322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,5120,0.008299555215570662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,3584,0.005984000033802456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,4096,0.008025777836640676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,5120,0.018559111489189994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,4096,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,3072,0.005316444569163852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,3584,0.007836444510353936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,4096,0.016356445021099515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,8192,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,2560,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,3072,0.0075777777367168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,3072,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,3584,0.015818667080667283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,2048,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,2560,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,2560,0.006667555620272954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,3072,0.01443911095460256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,1536,0.004011555678314633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,2048,0.006326222171386083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,2048,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,2560,0.013375110924243927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,1536,0.005984000033802456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,51200,0.10730489095052083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,1536,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,1024,0.0031946665710873077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,1024,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,768,0.002797333316670524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,10240,0.011759111450778114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,1024,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,3584,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,512,0.0024622221373849446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,768,0.006019555446174409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,768,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,512,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,768,0.010684444672531553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,256,0.002161777681774563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,512,0.009646221995353699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,256,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,128,0.0021226666867733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,256,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,256,0.009302222066455418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,2048,0.013059555656380124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,64,0.001800000046690305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,128,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,1536,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,512,32,0.0018968888454967076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,65536,0.018087999688254464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,128,0.007280888656775157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,64,0.00499288903342353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,51200,0.015805333852767944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,1024,0.01053244454993142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,32,0.005104889058404499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,65536,0.04691288868586222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,65536,0.04879644513130188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,16384,0.007704888780911763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,51200,0.03770400087038676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,512,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,12288,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,65536,0.1345777776506212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,16384,0.01551555593808492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,51200,0.10702666971418594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,10240,0.006523555351628198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,16384,0.04004355602794223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,12288,0.013426666458447775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,12288,0.012708444562223224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,8192,0.005881777654091517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,10240,0.012071111136012606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,12288,0.031890667147106595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,10240,0.011365333365069495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,128,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,7168,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,8192,0.010067555639478896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,8192,0.010724444356229572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,10240,0.02812444501452976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,6144,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,7168,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,7168,0.009653333160612319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,5120,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,6144,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,6144,0.008999110923873054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,7168,0.022276444567574397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,4096,0.0058853332367208265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,51200,0.039111998346116804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,6144,0.020259555843141343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,5120,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,5120,0.008101333346631791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,3584,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,16384,0.015843555331230164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,4096,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,4096,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,3072,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,5120,0.018252443936136033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,3584,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,2560,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,4096,0.01717422240310245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,3072,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,3584,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,2048,0.004232888834344016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,2560,0.007191999918884701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,3072,0.014146667387750415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,2048,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,2560,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,8192,0.024783111280865137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,1536,0.0035546666218174826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,2048,0.012506666282812754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,1536,0.006565333240562015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,1024,0.0030586665703190696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,1536,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,1024,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,1024,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,3584,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,1024,0.010679999987284342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,768,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,3072,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,512,0.0024684444069862366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,768,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,2560,0.006680000159475539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,256,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,512,0.005676444619894028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,768,0.010024000373151567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,512,0.005062222066852781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,2048,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,128,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,256,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,512,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,128,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,64,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,128,0.004741333425045013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,32,0.0018568889548381169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,1536,0.012111999922328524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,128,0.007287110719415877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,256,768,0.0028008888992998335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,65536,0.011381333072980246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,64,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,32,0.004592000196377437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,51200,0.01034044474363327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,65536,0.046669334173202515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,65536,0.04655288987689548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,16384,0.006068444500366847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,51200,0.0373582210805681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,51200,0.03717777795261807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,65536,0.13394044505225286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,12288,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,16384,0.015840000576443143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,16384,0.015275556180212231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,51200,0.10601955652236938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,10240,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,256,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,12288,0.01311911145846049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,16384,0.04011111127005683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,8192,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,256,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,10240,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,12288,0.031896889209747314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,7168,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,8192,0.010660444696744284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,8192,0.01033866653839747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,10240,0.027800000376171533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,6144,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,7168,0.010069333016872406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,7168,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,6144,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,5120,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,6144,0.020682666036817763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,5120,0.008236444658703273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,12288,0.0129439996348487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,5120,0.008635555704434713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,4096,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,10240,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,3584,0.004972444640265571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,4096,0.008295111358165741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,5120,0.01890222231547038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,4096,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,3072,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,3584,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,3584,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,4096,0.016528000434239704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,8192,0.024030221833123103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,2560,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,3072,0.0073626670572492815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,3584,0.015971556305885315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,2048,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,6144,0.009394666386975182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,2560,0.00702311098575592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,3072,0.014848889576064216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,1536,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,2048,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,2048,0.0064222220745351576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,1536,0.00645955569214291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,1536,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,1024,0.0028817777832349143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,1536,0.011770666473441653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,1024,0.0057751110030545135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,768,0.0028435554769304064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,1024,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,1024,0.011363555159833698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,768,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,3072,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,512,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,7168,0.022311111291249592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,2560,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,768,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,256,0.0021848889688650766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,512,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,512,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,128,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,256,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,256,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,2048,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,64,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,256,0.008973333570692275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,128,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,128,32,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,65536,0.008294222255547842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,64,0.004853333450025982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,128,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,32,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,16384,0.005706666658322017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,65536,0.04585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,12288,0.005112888912359874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,768,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,51200,0.03699466586112976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,10240,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,16384,0.015170666906568738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,8192,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,12288,0.012788444757461548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,2560,0.013431111143694984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,10240,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,6144,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,8192,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,512,0.01035911093155543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,5120,0.004621333132187526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,4096,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,7168,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,128,0.0047848886913723415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,3584,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,6144,0.008732444710201686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,3072,0.004599110947714912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,2560,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,5120,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,2048,0.004607999904288186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,51200,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,1536,0.003552000141806073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,2560,0.006456888798210356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,1024,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,2048,0.006315555423498154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,768,0.002672000063790215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,1536,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,512,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,1024,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,256,0.0021271111650599376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,7168,0.005065777649482091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,128,0.0018702221827374564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,512,0.004969777746333016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,64,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,256,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,64,32,0.0018044445249769424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,128,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,64,0.004612444589535396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,4096,0.007473777565691207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,32,0.004852444347408083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,51200,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,16384,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,3584,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,3072,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,65536,0.045893334680133395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,10240,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,51200,0.03700000047683716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,8192,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,12288,0.01256622208489312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,7168,0.005069333232111401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,6144,0.004851555658711327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,768,0.0052915554907586836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,5120,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,8192,0.010062221851613786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,4096,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,7168,0.009543110926946005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,3584,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,6144,0.00868622213602066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,3072,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,5120,0.008298666940795051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,2560,0.005106666849719154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,4096,0.007354666789372762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,2048,0.004239999999602635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,65536,0.008647110727098253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,1536,0.0035093331502543557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,3584,0.007245333658324347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,1024,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,3072,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,768,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,2560,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,512,0.002536000063021978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,2048,0.006171555568774541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,256,0.002128888956374592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,1536,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,128,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,768,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,16384,0.01516888870133294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,64,0.0018311111877361934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,512,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,32,0.0017928888814316855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,256,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,10240,0.011319110790888468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,128,0.004950222041871813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,64,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,32,0.005076444397370021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,16384,0.3843253453572591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,16384,0.19939733876122367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,16384,0.1318862173292372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,51200,0.5990764300028483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,80,32,12288,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,12288,0.2896044519212511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,1024,0.005501333210203383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,12288,0.14785422219170466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,10240,0.218213333023919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,51200,1.2066951327853732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,12288,0.099254224035475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,8192,0.17136977778540718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,7168,0.15298844708336726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,10240,0.08452533351050483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,8192,0.06744355625576444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,8192,0.1063457793659634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,51200,0.39292176564534503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,6144,0.13251733779907227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,7168,0.06073333157433403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,5120,0.11267999807993571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,6144,0.05225511060820686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,6144,0.0784426662656996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,5120,0.04542577928966946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,3584,0.07748533619774713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,4096,0.04448622133996752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,4096,0.054457777076297335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,5120,0.06575555271572538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,3072,0.0692302253511217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,10240,0.12494666046566433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,3584,0.034317334493001304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,3584,0.048250665267308555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,2560,0.058000889089372426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,3072,0.042167110575569995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,2048,0.04728177852100796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,2560,0.031090666850407917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,7168,0.09232266743977864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,2560,0.03675999906327989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,1536,0.03583111034499274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,1024,0.026334222820070054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,2048,0.02296888828277588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,2048,0.03144177794456482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,768,0.020187555087937247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,4096,0.09290399816301133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,1536,0.044215109613206655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,1536,0.01851466629240248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,1024,0.014846222268210517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,1024,0.03178933262825012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,256,0.008243555823961893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,768,0.02495288848876953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,768,0.01350933313369751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,512,0.012392889294359418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,128,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,512,0.01758399936887953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,256,0.010839111275143094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,64,0.004500444564554426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,128,0.010724444356229572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,256,0.012634666429625617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,128,0.011152889165613385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,3072,0.030223998758527968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,64,0.011361777782440186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,32,0.01551999979548984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,512,0.013478221992651621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,65536,0.6200044420030382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,51200,0.4710639847649469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,51200,0.8505751291910807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,16384,0.2755253314971924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,65536,1.09443113538954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,32,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,12288,0.208259556028578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,16384,0.1582453383339776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,10240,0.17320888572269014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,16384,0.11121866438123916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,12288,0.1178284486134847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,65536,0.41716978285047746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,8192,0.1409297784169515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,10240,0.07093955410851373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,51200,0.3223671118418376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,8192,0.05577422512902153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,7168,0.12467822763654922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,10240,0.09975377718607585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,8192,0.08501599894629584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,6144,0.10741511318418716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,5120,0.09045688973532783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,7168,0.06202666627036201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,6144,0.06332266992992826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,4096,0.07382577657699585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,5120,0.03908355699645148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,6144,0.04359555575582716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,3584,0.06524888674418132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,3072,0.05677600039376152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,4096,0.03176444437768724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,4096,0.04412710997793409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,2560,0.047726223866144814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,3072,0.024912888805071514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,3584,0.03963288995954726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,2048,0.03882666760020786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,2560,0.025915554828113977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,2560,0.030037333567937214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,1536,0.030653334326214258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,2048,0.026163554853863184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,2048,0.019327999817000497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,1024,0.022553778356975977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,7168,0.07182044453091092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,1536,0.015496888094478183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,768,0.016215999921162922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,1024,0.013071111506885953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,1024,0.02609866691960229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,5120,0.05310488740603129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,512,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,1536,0.03764444589614868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,256,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,768,0.01998488936159346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,512,0.010798222488827176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,128,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,512,0.01517599986659156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,3584,0.028387556473414104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,256,0.009686222506894005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,64,0.0038737778862317405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,256,0.010604444477293225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,3072,0.0343537794219123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,128,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,128,0.009728888670603434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,64,0.010065777434243096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,32,0.01311911145846049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,65536,0.31681421067979604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,51200,0.24958933724297416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,65536,0.22316977712843153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,16384,0.08230311340755887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,65536,0.14361955059899226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,768,0.012028444144460889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,51200,0.17082666026221383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,12288,0.06310222546259563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,16384,0.04035733474625482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,16384,0.05933244360817803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,12288,0.045364442798826426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,51200,0.11839289135403103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,12288,0.03303822212749057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,10240,0.05293777916166517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,32,0.003957333250178231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,8192,0.04514400164286295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,7168,0.03826577795876397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,10240,0.027982221709357366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,8192,0.023359111613697473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,12288,0.0911137792799208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,6144,0.03331822156906128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,8192,0.03404622276624044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,7168,0.026240888569090102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,5120,0.02864444586965773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,10240,0.039111998346116804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,7168,0.029450668228997126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,4096,0.024255999260478552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,3584,0.02071199980047014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,6144,0.019474666979577806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,6144,0.02629866699377696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,5120,0.017231111725171406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,3072,0.01830844415558709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,3584,0.014074666632546319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,2560,0.015406222807036506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,5120,0.022618666291236877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,4096,0.019174221489164565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,3584,0.017489777670966256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,2048,0.012616000241703458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,3072,0.013172444370057849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,1536,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,2560,0.013730666703648038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,3072,0.015525332755512662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,1024,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,2048,0.010760889285140567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,2560,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,768,0.006344888773229387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,2048,0.01239911135700014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,1536,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,512,0.004834666848182678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,1536,0.015464888678656684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,1024,0.00775111135509279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,256,0.003156444471743372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,768,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,1024,0.012177777787049612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,128,0.0028008888992998335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,512,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,768,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,64,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,256,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,32,0.002536888958679305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,128,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,256,0.00654755574133661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,128,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,64,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,32,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,65536,0.2503093348609077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,51200,0.19520621829562715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,16384,0.0682800014813741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,65536,0.1706311172909207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,4096,0.015138667490747241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,65536,0.12378488646613227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,51200,0.1310453282462226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,12288,0.05428266525268555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,51200,0.10014666451348199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,10240,0.04451644420623779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,16384,0.03697333402103848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,8192,0.03787733448876275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,12288,0.031035555733574763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,7168,0.032538665665520564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,10240,0.025428444147109985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,10240,0.03120622369978163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,8192,0.021526222427686054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,6144,0.028060442871517603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,8192,0.027104889353116352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,5120,0.024085332949956257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,512,0.008063111040327284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,6144,0.01772533357143402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,4096,0.020635555187861126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,6144,0.021204445097181533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,5120,0.015832000308566622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,5120,0.017895110779338412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,3584,0.0177875558535258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,4096,0.01425866617096795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,4096,0.015558222929636637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,3072,0.01571111049917009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,3584,0.013406222065289816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,3584,0.01420177850458357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,2560,0.013064888616402945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,3072,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,3072,0.013008000122176277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,16384,0.046776887443330556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,12288,0.03593066665861342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,2560,0.012418666647540199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,1536,0.009273777405420939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,2560,0.011824888487656912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,2048,0.009604444106419882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,1024,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,2048,0.011071110765139261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,7168,0.02273333403799269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,768,0.006030222194062338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,7168,0.023687111006842718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,1536,0.013070222404268054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,1024,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,512,0.004548444516128964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,1024,0.010017777482668558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,768,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,256,0.0031928889867332247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,512,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,128,0.0028311111446883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,768,0.008405333591832055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,512,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,256,0.006607111129495833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,64,0.002549333290921317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,32,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,256,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,128,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,64,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,32,0.0070266665683852295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,65536,0.2209573321872287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,51200,0.1746764447953966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,65536,0.16216888692643908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,16384,0.06323911084069146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,65536,0.1196240054236518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,1536,0.008363555702898238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,51200,0.12553866704305014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,12288,0.04895022180345324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,10240,0.04069777660899692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,12288,0.027664000789324444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,12288,0.034295111894607544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,8192,0.034281777011023626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,10240,0.023970666858885024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,10240,0.029951112137900457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,128,0.006362666686375936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,7168,0.03003466791576809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,8192,0.025986666480700176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,8192,0.02142133315404256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,6144,0.025984888275464375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,7168,0.019263111882739596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,2048,0.011165333290894827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,7168,0.022455111145973206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,6144,0.01720266706413693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,4096,0.019066666563351948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,6144,0.019908444748984445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,5120,0.015602666470739575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,5120,0.017268444101015728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,4096,0.013827555709415011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,51200,0.09888711240556504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,3072,0.014088888963063559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,16384,0.044700443744659424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,3584,0.012982221941153208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,3584,0.013999111122555204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,2560,0.012392889294359418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,3072,0.012111999922328524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,2048,0.010671111444632212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,3072,0.012893333203262754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,2560,0.011379555695586734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,1536,0.009191999832789103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,2560,0.01163822246922387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,2048,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,1024,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,2048,0.010663111176755695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,1536,0.008064889245563084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,768,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,1536,0.01145600030819575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,5120,0.02215555475817786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,1024,0.007325333025720384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,512,0.004452444612979889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,1024,0.008859555754396651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,768,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,256,0.003370666669474708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,768,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,512,0.0068408888247278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,3584,0.0173342227935791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,256,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,512,0.006763555523422029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,64,0.0026364444444576898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,256,0.006100444330109491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,32,0.0028204443968004654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,128,0.00630666646692488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,4096,0.01517688896920946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,128,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,16384,0.034862223598692156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,64,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,65536,0.1739413340886434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,32,0.006519111080302133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,51200,0.13199467129177517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,65536,0.11928889486524795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,65536,0.07000888718499078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,12288,0.03652000096109178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,51200,0.05680977635913425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,16384,0.022625777456495497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,10240,0.030200001266267564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,16384,0.034437331888410784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,12288,0.018232888645595975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,12288,0.026535999443795946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,10240,0.016207999653286405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,128,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,7168,0.022317333353890315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,8192,0.014479999740918478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,8192,0.0199306673473782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,6144,0.01936533384852939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,7168,0.013429332938459186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,7168,0.017619555195172627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,5120,0.016608888904253643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,6144,0.012444444828563266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,16384,0.04895377821392483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,4096,0.014119999276267158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,5120,0.011554666691356234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,5120,0.014017777310477363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,3584,0.012296888563368054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,4096,0.010592000352011787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,3072,0.011044444309340583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,4096,0.0127697777416971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,3584,0.009491555392742157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,2560,0.009485333330101436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,8192,0.025732444392310247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,3584,0.011720000041855706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,2048,0.008328888979223039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,3072,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,10240,0.023008000519540574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,2560,0.008778666456540426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,1536,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,2048,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,2560,0.009986666341622671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,1024,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,2048,0.008915555146005418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,1536,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,1536,0.00977600034740236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,1024,0.006425777657164468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,512,0.0034533333447244433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,1024,0.008042666647169325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,768,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,6144,0.015803555647532146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,256,0.0026497778793176017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,512,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,128,0.00249422217408816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,512,0.0063279999627007385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,256,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,51200,0.09218844440248276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,64,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,32,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,256,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,128,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,128,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,64,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,65536,0.15520711739857992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,32,0.006227555374304454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,3072,0.010820444259378644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,51200,0.12458844979604085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,65536,0.11618222130669488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,65536,0.06675200329886542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,12288,0.03433422247568766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,51200,0.05477244324154324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,768,0.004231111043029361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,51200,0.08948889043596055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,16384,0.02166044380929735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,10240,0.028402666250864666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,16384,0.033168001307381526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,8192,0.02442488902144962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,12288,0.017824888229370117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,12288,0.025591999292373657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,7168,0.021162667208247717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,768,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,10240,0.022327999273935955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,6144,0.01853511068556044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,8192,0.019367999500698514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,7168,0.013050666285885705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,5120,0.01573866605758667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,7168,0.016931555337376065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,6144,0.012149333126015134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,4096,0.013910222384664746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,6144,0.015490666031837463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,5120,0.011017777853541903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,4096,0.009753777749008602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,4096,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,3072,0.010653333531485664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,3584,0.008738666772842407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,3584,0.011563555234008364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,16384,0.04455022348297966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,2560,0.009072889056470659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,3072,0.00830844458606508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,3072,0.010776888993051318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,2048,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,2560,0.007673777639865875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,2560,0.00980444418059455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,1536,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,10240,0.015838222371207345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,2048,0.007115555306275685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,8192,0.014150222142537435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,1024,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,2048,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,1536,0.007028444773621029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,768,0.003915555361244413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,1536,0.009974222216341231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,1024,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,512,0.0032239999208185407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,768,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,1024,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,256,0.002586666701568498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,3584,0.011694221860832639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,512,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,768,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,128,0.0022408889813555614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,5120,0.013797333670987023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,64,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,256,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,512,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,128,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,128,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,65536,0.12387022707197402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,64,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,32,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,51200,0.09889155626296997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,16384,0.03716622127426995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,65536,0.06514488988452487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,65536,0.11108533541361491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,12288,0.03086577852567037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,51200,0.052878220876057945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,51200,0.08564177486631606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,10240,0.026904889278941687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,12288,0.017256889078352187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,8192,0.024131556351979572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,12288,0.024485333098305598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,10240,0.015238222148683337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,10240,0.02132355504565769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,7168,0.01960177719593048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,8192,0.013787555197874704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,6144,0.01703555550840166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,7168,0.012714666624863943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,7168,0.0165928883685006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,5120,0.014818666709793938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,6144,0.011685333318180509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,6144,0.015159999330838522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,4096,0.012670222255918713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,5120,0.010598222414652506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,5120,0.013577777478430005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,3584,0.010671111444632212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,4096,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,4096,0.012048000262843238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,3072,0.010228444304731157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,3584,0.008393777741326226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,16384,0.021193777521451313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,2560,0.008628444539176093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,3584,0.01108266661564509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,3072,0.00793333351612091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,2048,0.008956444760163626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,3072,0.01037333326207267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,2560,0.0075839997993575195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,1536,0.006450666735569636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,2048,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,2560,0.009275555610656738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,1024,0.0047048889100551605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,2048,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,1536,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,8192,0.018587556150224473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,768,0.003839111162556542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,512,0.0032791110376516977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,1536,0.009080000221729279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,1024,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,32,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,256,0.0026142222599850762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,768,0.006002666635645761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,1024,0.007442666424645319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,128,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,512,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,768,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,256,0.00574933323595259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,512,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,128,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,32,0.0023173333870040043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,128,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,65536,0.10843288898468018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,64,0.005391110976537068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,32,0.006022222340106964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,51200,0.08744621939129299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,65536,0.10590310891469319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,16384,0.03152444296412998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,65536,0.06284088558620878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,12288,0.027817779117160376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,51200,0.050422221422195435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,51200,0.0832915571000841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,16384,0.020252444677882724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,10240,0.023249778482649062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,16384,0.030361778206295435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,12288,0.016897777716318767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,8192,0.020774222082561918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,12288,0.023590222001075745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,10240,0.020662221643659804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,7168,0.018171555466122098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,8192,0.013229333692126803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,64,0.0021271111650599376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,8192,0.018226666582955253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,6144,0.01577333278126187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,7168,0.012378666963842181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,5120,0.013632000320487551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,7168,0.0162444445821974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,6144,0.011076444553004371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,4096,0.010694222317801582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,6144,0.014874666929244995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,5120,0.010032888915803697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,3584,0.009269333548016017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,16384,0.03224533465173509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,5120,0.013258666627936892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,4096,0.008692444198661381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,3072,0.008645333349704742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,4096,0.011731555892361535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,2560,0.008578666382365758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,3584,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,3072,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,3584,0.010983111129866706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,2048,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,3072,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,1536,0.005981333139869902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,2560,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,10240,0.014893333117167154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,2048,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,2560,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,1024,0.004559111264016894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,768,0.003846222327815162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,1536,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,1024,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,512,0.0034906665484110513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,768,0.005980444451173146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,1024,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,256,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,768,0.006315555423498154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,256,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,512,0.005976888868543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,128,0.002472888885272874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,512,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,256,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,64,0.0021520000365045336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,32,0.0021688888470331826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,256,0.005499555418888728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,128,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,128,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,65536,0.08601333035363092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,64,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,51200,0.06979644298553467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,32,0.005679111099905438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,16384,0.028756442997190688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,65536,0.060405333836873375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,65536,0.06464355521731906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,12288,0.022267555197079975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,51200,0.050981332858403526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,16384,0.019555555449591745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,10240,0.019581332802772522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,2048,0.00795199970404307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,16384,0.019613333874278598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,8192,0.01718133356836107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,12288,0.016249777542220224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,1536,0.008045333127180735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,12288,0.01585422290696038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,7168,0.015134221977657743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,10240,0.01443733274936676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,6144,0.013778666655222574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,8192,0.012682666381200155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,5120,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,8192,0.013292444248994192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,6144,0.010639111200968424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,7168,0.011753777662913004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,4096,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,6144,0.011028444601429833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,5120,0.009188444250159794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,5120,0.010025777750545079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,4096,0.008642666869693333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,3072,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,4096,0.009004444711738164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,3584,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,3584,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,2560,0.006508444332414203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,3072,0.007736889024575551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,2048,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,3072,0.007496889266702864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,2560,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,51200,0.048519111341900296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,1536,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,2560,0.0068986668354935115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,2048,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,1024,0.003559999995761447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,1536,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,2048,0.006653333289755716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,768,0.003206222214632564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,1536,0.007249777515729268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,1024,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,10240,0.014436443646748861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,512,0.0028640000770489373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,1024,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,768,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,7168,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,256,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,768,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,128,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,256,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,3584,0.009347555538018545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,64,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,128,0.005868444426192178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,32,0.0023164444913466773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,256,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,128,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,65536,0.08362755510542129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,64,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,51200,0.06680622365739611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,32,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,16384,0.025851555996470984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,65536,0.05938133266237047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,51200,0.04738755689726936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,12288,0.020136000381575692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,16384,0.01907911068863339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,12288,0.015847111741701763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,10240,0.017246221502621967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,12288,0.015517334143320719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,512,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,8192,0.01555022266176012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,512,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,10240,0.014273777604103088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,7168,0.014096000128322177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,10240,0.01407999959256914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,8192,0.012467555701732635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,6144,0.013082666529549493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,7168,0.011132444772455426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,8192,0.012670222255918713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,5120,0.01146933353609509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,7168,0.01145866678820716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,6144,0.010250666903124915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,4096,0.010204444328943888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,65536,0.06270133124457465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,6144,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,5120,0.009011555876996782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,3584,0.008777778181764815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,3072,0.007181333170996771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,5120,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,4096,0.008415111237102086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,51200,0.0491866668065389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,3584,0.00794044468137953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,2560,0.006414222220579783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,4096,0.008595555192894405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,2048,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,3584,0.007733333441946242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,1536,0.0042364444169733255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,2560,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,3072,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,1024,0.0034773333205117118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,2048,0.00721155520942476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,2560,0.006907555378145642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,2048,0.0064497776329517365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,1536,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,768,0.0031555555760860443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,512,0.002749333365095986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,1024,0.006230222268237009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,1536,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,256,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,768,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,512,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,768,0.005879999862776862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,128,0.0021253333737452826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,256,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,512,0.0054551110499435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,64,0.0021342221233579847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,256,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,32,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,128,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,65536,0.07335022423002455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,128,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,64,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,51200,0.05862400266859266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,32,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,16384,0.020787555310461257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,65536,0.05805333455403646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,65536,0.059812446435292564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,12288,0.017058667209413316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,3072,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,51200,0.046670221620135836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,51200,0.04732622371779548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,16384,0.018983999888102215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,10240,0.014579556054539151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,16384,0.018216888109842937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,8192,0.012542222109105853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,12288,0.015798222687509324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,10240,0.013926222920417786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,12288,0.015119999647140503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,7168,0.011873777541849347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,8192,0.011172444456153445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,6144,0.0107360002067354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,1024,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,7168,0.011015111373530494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,5120,0.009942221972677443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,7168,0.011029333704047732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,6144,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,4096,0.008451555338170793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,6144,0.01037333326207267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,5120,0.008991999758614434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,5120,0.009369778136412302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,3584,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,4096,0.008292444050312042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,3072,0.0069066666894488865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,4096,0.00795022232664956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,3584,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,2560,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,16384,0.01926577753490872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,3584,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,3072,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,2048,0.0053297777970631914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,3072,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,2560,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,1536,0.0041680000722408295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,2048,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,1024,0.003495111233658261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,1536,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,2048,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,768,0.0029288888391521242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,10240,0.013502221968438892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,1536,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,512,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,1024,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,8192,0.012292444705963135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,256,0.0024986666523747975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,1024,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,768,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,128,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,512,0.005655999812814925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,768,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,256,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,64,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,512,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,32,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,128,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,256,0.005068444543414646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,65536,0.06291911337110731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,51200,0.04641777939266629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,64,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,32,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,16384,0.01885955532391866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,65536,0.05786666605207655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,51200,0.04552888870239258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,12288,0.01496355566713545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,16384,0.018813333577579923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,51200,0.045610666275024414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,10240,0.013440888788965015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,16384,0.017886221408843994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,2560,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,8192,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,10240,0.013409777647919126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,12288,0.014736000034544202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,7168,0.011022222538789114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,10240,0.013063111239009433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,6144,0.00996800015370051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,7168,0.010012444522645738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,7168,0.011026666396194033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,5120,0.009332444104883406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,6144,0.009656888743241629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,6144,0.009940444595283931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,4096,0.008096888661384583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,5120,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,5120,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,3584,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,4096,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,65536,0.056872891055213086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,4096,0.00795911086930169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,3072,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,2560,0.0058622223635514574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,3584,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,3072,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,3584,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,2048,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,2560,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,3072,0.007185777856243982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,12288,0.015204444527626038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,1536,0.0038568890757030914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,2560,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,1024,0.0032382222513357797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,2048,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,768,0.0028248888750871024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,1536,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,8192,0.011176889141400656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,2048,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,1024,0.006261333409282897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,512,0.002512000087234709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,1536,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,8192,0.011872889267073737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,256,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,1024,0.005881777654091517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,768,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,128,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,512,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,768,0.005676444619894028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,64,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,256,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,32,0.002151111140847206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,128,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,65536,0.050423112180497914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,128,0.004944889081848992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,64,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,32,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,65536,0.05564710828993055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,16384,0.016499555773205228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,65536,0.05471200082037184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,51200,0.04493244489034017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,51200,0.04392533169852363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,16384,0.018331555856598746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,16384,0.01756977703836229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,10240,0.011892444557613797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,12288,0.014825777875052558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,12288,0.014554666148291694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,8192,0.010666666759385003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,10240,0.012656889028019376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,10240,0.013075555364290873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,7168,0.009966221948464712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,8192,0.011086222198274402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,512,0.005425777700212266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,6144,0.010284444524182213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,256,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,7168,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,5120,0.009009777671760982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,7168,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,51200,0.04038044479158189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,4096,0.009714666340086196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,6144,0.009368889033794403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,6144,0.009277332988050249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,5120,0.008927110996511247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,3584,0.011360000405046674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,5120,0.008406221866607666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,4096,0.008310221963458592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,3072,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,3584,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,2560,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,3072,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,3072,0.007024889190991719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,2048,0.006029333505365584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,2560,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,2560,0.006779555645253923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,1536,0.005160888863934411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,2048,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,2048,0.006343110981914733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,1024,0.0038817777401871155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,8192,0.011648889217111798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,768,0.003039999968475766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,1536,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,1024,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,512,0.0025902222841978073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,768,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,768,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,256,0.002346666736735238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,512,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,512,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,128,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,12288,0.01348088930050532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,256,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,3584,0.0075839997993575195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,128,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,32,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,128,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,65536,0.038853334056006536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,64,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,32,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,1536,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,51200,0.031152000029881794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,65536,0.05314400129848056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,16384,0.013468444347381592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,1024,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,51200,0.044459554884168834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,12288,0.013728888498412238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,51200,0.04194044404559665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,16384,0.017525333497259352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,16384,0.017139555679427255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,10240,0.012430222498046027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,12288,0.013422222601042854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,8192,0.011005333728260465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,4096,0.007685333490371704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,12288,0.014104000396198697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,10240,0.012098666694429187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,7168,0.010400888820489248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,64,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,6144,0.00958933350112703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,8192,0.010724444356229572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,7168,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,8192,0.01038755559258991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,5120,0.008993777963850234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,7168,0.009752888646390703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,6144,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,4096,0.007870222131411234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,6144,0.009104000197516548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,5120,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,3584,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,5120,0.0083137775460879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,4096,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,3072,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,4096,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,3584,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,2560,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,3584,0.007847111258241866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,3072,0.0075902218619982404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,2048,0.00453599997692638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,3072,0.007045333584149678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,2560,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,1536,0.004578666554556953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,2560,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,2048,0.0069146665434042616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,1024,0.0031395554542541504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,2048,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,1536,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,256,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,768,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,1536,0.006176888942718506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,512,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,1024,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,1024,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,768,0.005886222339338726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,256,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,512,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,768,0.005542222410440445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,128,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,256,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,64,0.0017706666969590718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,128,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,256,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,65536,0.054775999652014844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,32,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,65536,0.028485334581798975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,128,0.00508622204264005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,51200,0.022861333356963262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,64,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,32,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,65536,0.054414222637812294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,16384,0.011692444483439127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,65536,0.05068177647060818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,51200,0.04400888747639126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,12288,0.010822222464614444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,51200,0.04043733411365085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,16384,0.015836444165971544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,16384,0.01651733285850949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,12288,0.01296622223324246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,8192,0.009207111265924241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,10240,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,10240,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,7168,0.01036177741156684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,8192,0.01073688848151101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,512,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,6144,0.009375111096435124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,7168,0.010048000348938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,7168,0.009713778065310584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,5120,0.00944266633854972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,6144,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,4096,0.007701333198282454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,6144,0.009009777671760982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,5120,0.008820444345474243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,3584,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,5120,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,4096,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,3072,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,4096,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,3584,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,2560,0.004975999808973736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,10240,0.010217777556843227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,3072,0.007612444460391998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,3072,0.0070257774657673305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,2048,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,10240,0.012647111382749347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,2560,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,1536,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,2560,0.006565333240562015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,2048,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,1024,0.0038311113086011675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,1536,0.006298666612969504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,2048,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,768,0.003536888708670934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,1536,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,8192,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,512,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,1024,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,768,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,256,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,768,0.005440000030729506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,128,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,512,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,256,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,64,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,256,0.005020444591840108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,128,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,128,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,64,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,65536,0.022805333137512207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,12288,0.013217777841620974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,32,0.004960888789759742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,51200,0.02257866660753886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,65536,0.05429688758320279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,65536,0.049545778168572314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,16384,0.011013333168294696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,1024,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,51200,0.04349689020050896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,51200,0.040023999081717596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,12288,0.009407999614874521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,16384,0.015812445018026564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,512,0.00574311117331187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,10240,0.008950221869680617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,12288,0.013230221966902414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,16384,0.01536177761024899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,8192,0.008782222039169734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,12288,0.012784000072214337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,7168,0.008664888640244802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,10240,0.012075555821259817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,8192,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,10240,0.011714666253990598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,32,0.0021484444538752237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,8192,0.01033155537313885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,7168,0.010030222435792288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,5120,0.008151111503442129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,7168,0.009645333720578088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,3584,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,4096,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,6144,0.00941066692272822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,3584,0.006216889040337668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,5120,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,6144,0.009076444639099969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,3072,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,4096,0.008015111088752747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,5120,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,2560,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,4096,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,3072,0.007325333025720384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,2048,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,3072,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,2560,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,1536,0.0038657776183552216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,2048,0.006667555620272954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,2560,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,1024,0.003212444484233856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,2048,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,1536,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,768,0.002999111182159848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,1536,0.00611200018061532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,512,0.0025128889828920364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,768,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,1024,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,256,0.002181333386235767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,768,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,512,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,128,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,6144,0.009287999735938178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,256,0.005431111074156231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,512,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,768,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,256,0.004965333475006951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,128,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,65536,0.020612445142534044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,51200,0.017172444197866652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,128,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,64,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,65536,0.05336533321274651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,32,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,16384,0.011794666449228922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,3584,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,51200,0.03871377640300327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,65536,0.048600001467598804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,10240,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,3584,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,51200,0.03933777742915683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,8192,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,12288,0.013109332985348172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,16384,0.01519377695189582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,10240,0.012057777908113269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,12288,0.012721777790122561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,8192,0.01069155583779017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,10240,0.01144266708029641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,6144,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,8192,0.010225777824719747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,7168,0.010004444254769219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,5120,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,7168,0.009443555441167619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,6144,0.009365333451165093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,4096,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,5120,0.008771555291281806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,6144,0.00868622213602066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,3584,0.005976000179847081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,5120,0.008282666405042013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,4096,0.008262222011884054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,3072,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,4096,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,3584,0.007276444799370236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,2560,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,3072,0.007364444434642792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,3072,0.0069919998447100324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,2048,0.004546666724814309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,12288,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,2560,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,16384,0.015802666544914246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,2560,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,2048,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,1024,0.0032044444233179092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,7168,0.006665777828958299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,1536,0.00655644428398874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,768,0.0028364445186323593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,1536,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,1024,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,1024,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,512,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,768,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,256,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,768,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,512,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,128,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,512,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,256,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,64,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,256,0.005019555489222209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,32,0.0018293333964215384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,128,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,65536,0.018015111486117046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,64,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,32,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,51200,0.015792888071801927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,512,1536,0.0038862224254343244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,65536,0.04759200082884895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,65536,0.04654755526118808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,16384,0.007830222447713217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,2048,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,51200,0.037440889411502414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,16384,0.015817777978049386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,12288,0.007631111476156447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,51200,0.03762933280732896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,10240,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,16384,0.015300444430775113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,12288,0.01311911145846049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,10240,0.011731555892361535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,12288,0.012702222499582501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,7168,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,10240,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,8192,0.0105253333846728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,8192,0.010030222435792288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,6144,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,7168,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,3584,0.00795999997191959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,5120,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,7168,0.009537777966923183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,128,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,4096,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,6144,0.009305777649084726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,3584,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,6144,0.008790222307046255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,5120,0.008656000097592672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,3072,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,4096,0.008011555506123437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,2560,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,3584,0.007703999678293864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,4096,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,3072,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,3584,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,2560,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,3072,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,1536,0.003556444413132138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,2560,0.006572444405820634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,1024,0.0029324444217814338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,2048,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,1536,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,8192,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,768,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,1536,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,512,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,1024,0.005983110931184556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,1024,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,256,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,512,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,768,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,128,0.0018737777653667664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,512,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,64,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,256,0.004868444469239977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,128,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,32,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,65536,0.011362666885058085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,128,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,5120,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,51200,0.012373333175977072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,64,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,65536,0.045941332976023354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,256,2048,0.0041928887367248535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,16384,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,65536,0.04519733455446032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,51200,0.03734933336575826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,2048,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,12288,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,16384,0.015860444969601102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,10240,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,16384,0.015273777974976433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,12288,0.0131217779384719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,8192,0.005738666488064661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,12288,0.0125608891248703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,768,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,7168,0.005542222410440445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,10240,0.011350221931934357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,8192,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,6144,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,8192,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,7168,0.00999466660949919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,5120,0.006340444501903322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,6144,0.009334222310119206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,6144,0.008743999732865227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,4096,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,5120,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,5120,0.008101333346631791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,32,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,3584,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,4096,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,3072,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,3584,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,51200,0.036869333850012884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,2560,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,3584,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,3072,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,2048,0.004201777693298128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,3072,0.0068408888247278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,2560,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,1536,0.0035297779573334586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,10240,0.012067555553383298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,2560,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,1024,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,2048,0.006649777707126405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,2048,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,768,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,1024,0.006224000205596288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,7168,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,1024,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,768,0.005970666805903117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,768,0.0052266667286554975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,256,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,512,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,512,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,128,0.001863111017478837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,256,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,256,0.004860444615284602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,4096,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,64,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,128,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,32,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,65536,0.012738666600651212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,128,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,51200,0.011362666885058085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,64,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,16384,0.00564533347884814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,32,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,12288,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,10240,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,51200,0.0366248885790507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,8192,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,1536,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,7168,0.005208000126812193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,128,512,0.002512000087234709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,6144,0.00564177789621883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,10240,0.011067555182509951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,5120,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,8192,0.010016000105275048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,4096,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,7168,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,3584,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,6144,0.00869777798652649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,3072,0.004567999806669024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,5120,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,2560,0.004603555632962121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,4096,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,2048,0.004575110971927643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,3584,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,1536,0.0034862222770849862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,3072,0.006990222467316522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,1024,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,2560,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,768,0.002551111082235972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,2048,0.006247111078765657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,512,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,1536,0.00563733321097162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,65536,0.0451564457681444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,256,0.0021173333128293357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,1536,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,1024,0.0053084443012873335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,64,0.0018195555441909367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,768,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,16384,0.014825777875052558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,32,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,256,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,65536,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,64,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,51200,0.012804444465372296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,32,0.004553777890072929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,16384,0.007000000112586551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,65536,0.045202665858798556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,12288,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,51200,0.03660444418589274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,10240,0.005381333331267039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,16384,0.01479022204875946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,8192,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,12288,0.012426666915416718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,7168,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,10240,0.011182222101423474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,6144,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,64,128,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,8192,0.010011555420027839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,5120,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,4096,0.004876444323195351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,7168,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,3584,0.004846222284767362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,512,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,3072,0.004701333327425851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,128,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,5120,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,2560,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,4096,0.007589333587222629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,2048,0.004555555681387584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,1536,0.003540444291300244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,3584,0.007105777661005656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,1024,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,3072,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,768,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,2560,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,512,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,2048,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,256,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,1536,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,128,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,1024,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,64,0.0018648888087934917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,768,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,512,0.004944889081848992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,256,0.004857777721352047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,128,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,64,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,32,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,12288,0.012563555604881711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,16384,0.36366666687859434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,51200,0.49770222769843203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,16384,0.16742932796478271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,65,32,32,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,12288,0.2719759941101074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,16384,0.13725866211785212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,12288,0.1251466671625773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,6144,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,51200,0.4191768964131673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,12288,0.10365955697165595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,10240,0.23322044478522408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,8192,0.1864853302637736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,10240,0.105877333217197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,10240,0.08807644579145645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,51200,1.1270426644219291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,8192,0.08879644340938991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,8192,0.07144177622265287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,7168,0.162538660897149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,16384,0.28853599230448407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,12288,0.21523377630445692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,6144,0.1388017733891805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,7168,0.07556977536943224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,10240,0.18103733327653673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,6144,0.06624266836378309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,7168,0.06794310940636529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,8192,0.14421599441104466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,5120,0.11888266934288873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,6144,0.05522844526502821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,4096,0.09483999676174587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,5120,0.04662755462858412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,5120,0.056130667527516685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,4096,0.04622755448023478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,3584,0.08423466814888848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,7168,0.12939377625783285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,6144,0.11328266726599799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,4096,0.03900977638032701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,3072,0.07271911038292779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,5120,0.0942231085565355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,4096,0.0779306689898173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,3584,0.03534222311443753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,2560,0.061420447296566434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,3584,0.041555556986067034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,3072,0.03162044286727905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,3072,0.036658667855792575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,2048,0.05023466547330221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,2560,0.03171022070778741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,2560,0.02920711040496826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,1536,0.03923644291030036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,3072,0.06425511174731784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,3584,0.07191466622882418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,2048,0.027256889475716486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,2048,0.02328444520632426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,2560,0.05515022079149882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,1024,0.03022577696376377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,1536,0.018794667389657762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,1536,0.022056000100241765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,768,0.021943999661339655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,1024,0.014927110738224454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,1024,0.016519111063745286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,1536,0.05076800121201409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,2048,0.04880533284611172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,512,0.014356444279352823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,768,0.013236444857385425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,1024,0.03343200021319919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,768,0.013743110828929476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,256,0.008405333591832055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,512,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,512,0.010696888797812991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,128,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,768,0.028635554843478735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,256,0.008976000050703684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,256,0.010787555740939246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,512,0.025003555748197768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,128,0.01074399964676963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,32,0.003900444342030419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,128,0.00812266684240765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,64,0.009305777649084726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,128,0.01571822166442871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,32,0.012392000191741519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,51200,0.8648639784918891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,65536,0.5212240219116211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,64,0.003899555653333664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,65536,1.1060507032606337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,256,0.0207342223988639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,51200,0.8738320138719348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,16384,0.2822062174479167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,51200,0.3988835546705458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,65536,0.42994756168789333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,16384,0.1341573397318522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,12288,0.21164888805813262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,16384,0.11705066098107232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,12288,0.10099466641743977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,51200,0.33270756403605145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,12288,0.08757333623038398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,10240,0.17804711394839814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,10240,0.08580266767077976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,8192,0.14554844962226018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,10240,0.06944444444444445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,8192,0.07197866837183635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,16384,0.2303768793741862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,12288,0.1736595498190986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,8192,0.05693155527114868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,7168,0.12514577971564397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,6144,0.10758399963378906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,7168,0.055953780810038246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,7168,0.06184177928500705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,10240,0.14557154973347983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,8192,0.12022133668263753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,5120,0.09161511394712661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,6144,0.04366222355100843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,7168,0.10534933540556167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,5120,0.04593866732385424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,4096,0.0746213330162896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,5120,0.03760266635153029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,6144,0.09297066926956177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,3584,0.0654275549782647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,5120,0.08027999930911593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,4096,0.03820711043145921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,4096,0.03291822142071194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,3072,0.057542224725087486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,3584,0.03446311089727614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,3584,0.029595557186338637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,4096,0.0677493347062005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,3072,0.03054844339688619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,3584,0.06100444661246406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,2560,0.0484479996893141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,6144,0.05431911018159655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,2560,0.024265777733590867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,2048,0.03941511114438375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,3072,0.05520711011356778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,2048,0.019311111834314134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,2560,0.04857333170043098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,2048,0.022778666681713525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,1536,0.031038221385743883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,2048,0.04193955659866333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,1536,0.018234666850831773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,1024,0.02201244400607215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,51200,0.6999351183573405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,1536,0.03641244437959459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,768,0.016228444046444364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,1024,0.016216889023780823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,3072,0.02795377704832289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,1024,0.028933333026038274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,65536,0.9140692816840278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,512,0.011365333365069495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,768,0.011490667031870948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,768,0.012137778103351593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,256,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,2560,0.026711111267407734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,512,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,512,0.010649777948856354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,768,0.025543111893865798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,128,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,512,0.02130222154988183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,256,0.010031999813185798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,1536,0.0162471118900511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,256,0.018559111489189994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,64,0.0038577777643998465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,128,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,32,0.004193777839342753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,128,0.007368889119890001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,128,0.014160000615649752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,1024,0.013758222262064615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,64,0.00832622249921163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,32,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,65536,0.3471653461456299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,65536,0.18255644374423555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,51200,0.2715946568383111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,51200,0.13618489106496176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,16384,0.08168177472220527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,65536,0.15070933765835232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,51200,0.11944799953036839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,256,0.007696000238259633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,16384,0.04187377625041538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,16384,0.04944800006018745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,12288,0.037754668129815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,16384,0.05689777930577596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,10240,0.05237955517239041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,12288,0.03415911065207587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,10240,0.03242933419015672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,12288,0.04483111037148369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,10240,0.03159200151761373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,10240,0.03950044512748718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,51200,0.1444595522350735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,8192,0.027543110979927912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,7168,0.03810311026043362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,65536,0.18939377201928032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,8192,0.024530667397711012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,8192,0.033991999096340604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,6144,0.033257779147889875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,12288,0.06415288978152804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,7168,0.024392889605628118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,5120,0.02849510974354214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,6144,0.02204088866710663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,7168,0.031189332405726116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,6144,0.021976000732845728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,4096,0.023436443673239812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,5120,0.01904977858066559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,5120,0.01755555636352963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,6144,0.028584887584050495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,3584,0.020579554968410067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,4096,0.015169777803950839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,4096,0.01587733295228746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,8192,0.04552177919281853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,3072,0.018187556001875136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,3584,0.014628445108731588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,3584,0.014259555273585849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,4096,0.022496889034907024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,2560,0.01513777838812934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,3072,0.013421333498424955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,3584,0.020999999509917364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,2048,0.01239999963177575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,2560,0.011711111499203576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,2560,0.012431999875439538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,3072,0.018630221486091614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,1536,0.00963466697269016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,7168,0.023507555325826008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,2048,0.010816000401973724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,2560,0.017267554998397827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,2048,0.010811555716726514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,1024,0.007873777714040544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,1536,0.009448889229032729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,1536,0.009641778137948778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,768,0.006039111150635614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,2048,0.01702044407526652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,1024,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,1024,0.008011555506123437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,1536,0.014527999692493014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,512,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,5120,0.02527644402450985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,768,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,768,0.007087111473083496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,256,0.003173333282272021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,512,0.007159999675220913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,512,0.005877333382765452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,128,0.0028213332924577924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,256,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,3072,0.013282666603724161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,512,0.010601777997281818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,256,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,64,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,32,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,128,0.006671999891599019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,128,0.005294222384691238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,256,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,64,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,32,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,65536,0.2451866732703315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,65536,0.15531110763549805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,51200,0.194487108124627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,65536,0.13457422786288792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,51200,0.1183440022998386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,1024,0.013296888934241401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,16384,0.07073244121339586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,768,0.012094222009181976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,16384,0.042950222889582314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,16384,0.03764444589614868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,12288,0.04975377851062351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,16384,0.047842668162451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,65536,0.14901244640350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,128,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,12288,0.03287911083963182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,10240,0.04096533192528619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,12288,0.030452443493737116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,51200,0.1177484459347195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,12288,0.038175112671322294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,10240,0.02576711111598545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,10240,0.028536889288160536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,7168,0.029496000872717962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,8192,0.02157955533928341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,10240,0.03398844599723816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,8192,0.024785778588718835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,6144,0.025799999634424847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,7168,0.021016000045670405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,8192,0.03070755468474494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,51200,0.10480799939897324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,5120,0.022231111923853557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,7168,0.02754488918516371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,6144,0.018231110440360177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,6144,0.019578667150603402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,4096,0.01870577699608273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,5120,0.016207999653286405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,5120,0.01648533344268799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,6144,0.025010666913456384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,3584,0.016201777590645682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,4096,0.014485332700941296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,4096,0.014472888575659858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,5120,0.02257333364751604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,8192,0.03466044531928168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,3072,0.015135111080275642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,3584,0.01350933313369751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,3584,0.013462222284740873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,4096,0.01978755493958791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,2560,0.012488889197508493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,3072,0.012599999705950419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,3072,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,3584,0.018852444158660043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,2048,0.010670222342014313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,2560,0.011970666547616323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,7168,0.021692444880803425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,1536,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,2048,0.009747555686367882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,2048,0.010302222437328762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,2560,0.016861332787407767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,1024,0.00720355576939053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,1536,0.008613333106040955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,1536,0.008952889177534316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,2048,0.01573333309756385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,768,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,1024,0.0075937774446275495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,1536,0.014142221874660917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,1024,0.007149332927332983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,512,0.004538666870858935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,768,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,768,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,256,0.002977777686383989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,512,0.006807999892367258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,768,0.010664000279373594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,512,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,128,0.002536888958679305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,512,0.009666666388511658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,256,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,256,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,64,0.002521777732504739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,128,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,256,0.009345778160625035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,2560,0.011372444530328115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,64,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,128,0.0076044441925154785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,32,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,65536,0.2222364478641086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,51200,0.17407911353641084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,1024,0.012122666670216454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,65536,0.12640533182356092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,51200,0.11164355278015137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,51200,0.09957244661119248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,16384,0.053393777873780995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,128,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,32,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,12288,0.04121333360671997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,16384,0.03558933403756883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,65536,0.13746221860249838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,16384,0.04067733221583896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,10240,0.03422310948371887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,51200,0.11042933331595527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,12288,0.031361778577168785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,12288,0.02883111106024848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,16384,0.0447688897450765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,8192,0.028952889972262915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,65536,0.14529155360327825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,10240,0.024344889654053584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,10240,0.027385777897304956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,12288,0.0366195539633433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,7168,0.024882665938801233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,8192,0.023376888699001733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,8192,0.02083822256988949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,10240,0.0322053333123525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,7168,0.019887111253208585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,7168,0.020225778222084045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,8192,0.02885599931081136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,5120,0.018904889623324077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,7168,0.026377777258555096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,3072,0.017812444104088675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,5120,0.015476443701320224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,5120,0.015899555550681222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,4096,0.01609688831700219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,5120,0.021704889006084863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,4096,0.014269333746698169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,4096,0.013912889692518445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,3584,0.013401777380042605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,4096,0.019948444432682462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,6144,0.0218577782313029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,3584,0.013097777962684631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,3584,0.012867555850081973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,3072,0.012072888513406118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,3584,0.01834222177664439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,3072,0.012113778127564324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,2560,0.010093332992659675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,6144,0.017424000634087455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,3072,0.012094222009181976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,2048,0.008784888519181145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,2560,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,3072,0.017522666189405654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,2560,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,6144,0.024117334021462336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,1536,0.007120889094140794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,2048,0.009963555468453301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,2048,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,1024,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,2560,0.01608088943693373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,1536,0.008023999631404877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,768,0.004394666602214177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,2048,0.01568711135122511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,1024,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,1024,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,512,0.0034862222770849862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,768,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,1536,0.01347733371787601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,1024,0.011370666325092316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,768,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,256,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,512,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,768,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,128,0.00254133323000537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,256,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,6144,0.01809955636660258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,512,0.009650666680600908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,64,0.002453333387772242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,256,0.009687999884287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,128,0.005294222384691238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,32,0.002483555633160803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,128,0.006467555546098285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,65536,0.17504710621303984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,64,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,128,0.0075937774446275495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,32,0.006074666563007567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,1536,0.007663999994595845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,65536,0.09557955794864231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,51200,0.1412960025999281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,65536,0.07042844427956475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,16384,0.0486195551024543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,51200,0.07330222262276544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,51200,0.05736622545454237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,12288,0.03281955586539374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,512,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,16384,0.022973333795865376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,16384,0.02814488940768772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,10240,0.02739199995994568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,12288,0.018800000349680584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,16384,0.04387822084956699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,12288,0.02211911148495144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,65536,0.13622400495741102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,51200,0.10862133238050674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,10240,0.01937599976857503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,10240,0.016177778442700703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,12288,0.035880890157487653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,7168,0.02016799979739719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,8192,0.01482222146458096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,8192,0.01685244507259793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,6144,0.017816889617178176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,10240,0.03211466802491082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,7168,0.013752000199423896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,7168,0.014762666490342883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,8192,0.028055999014112685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,5120,0.015304888288180033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,7168,0.025252444876564875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,6144,0.013588444226317935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,4096,0.012772444221708508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,5120,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,6144,0.023694222172101338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,256,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,5120,0.02102577851878272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,4096,0.010761777559916178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,3584,0.012074666718641916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,4096,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,8192,0.023174222972657945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,3072,0.010761777559916178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,4096,0.01930488811598884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,3584,0.01038755559258991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,3584,0.009648888475365108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,2560,0.009315555294354757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,3072,0.00865955568022198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,3584,0.018017777138286166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,3072,0.009583111438486312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,2048,0.008186666501892937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,2560,0.008786666724416945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,3072,0.017311111092567444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,6144,0.013044444223244986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,1536,0.006870222174459034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,2560,0.016005333926942613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,2048,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,5120,0.011356444822417365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,1024,0.004967999955018361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,1536,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,2048,0.01479466590616438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,1536,0.006223111102978389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,768,0.0038364442686239877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,1024,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,1024,0.00565244464410676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,1536,0.013432888521088494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,512,0.0032186667538351486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,768,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,1024,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,512,0.0053297777970631914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,512,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,768,0.010031110710567897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,256,0.002512000087234709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,2560,0.008492444124486711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,128,0.002456888970401552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,256,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,512,0.009365333451165093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,256,0.004954666727119022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,64,0.0025031111306614345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,2048,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,128,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,32,0.00254133323000537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,256,0.009307555854320526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,128,0.005176888985766305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,65536,0.15403910477956137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,128,0.007812444534566667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,32,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,51200,0.12362489435407852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,65536,0.08684444427490234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,16384,0.04420888755056593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,768,0.006247111078765657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,65536,0.06693510876761542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,51200,0.06663999954859416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,51200,0.054551111327277295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,12288,0.03429244293106927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,16384,0.022227555513381958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,16384,0.02644355595111847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,10240,0.026034666432274714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,12288,0.020440000626775954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,65536,0.1353564394844903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,12288,0.017895110779338412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,8192,0.022080888350804646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,51200,0.10857244332631429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,12288,0.034952001439200506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,10240,0.015797333584891427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,10240,0.017861333158281114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,7168,0.019103111492262948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,8192,0.01587288909488254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,64,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,10240,0.03159200151761373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,6144,0.016901334126790363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,8192,0.027507555153634813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,7168,0.013797333670987023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,5120,0.014322666658295525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,6144,0.012068444656001197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,7168,0.024953777591387432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,6144,0.022852443986468848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,5120,0.011013333168294696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,5120,0.020627554919984605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,16384,0.04346755478117201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,4096,0.009755555954244401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,3584,0.01074488874938753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,8192,0.014152000347773233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,4096,0.018592889110247295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,7168,0.013068444199032254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,3584,0.009759999811649323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,3072,0.01032266683048672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,3584,0.009391110804345872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,6144,0.012995555169052549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,2560,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,3072,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,3584,0.017486222916179232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,3072,0.009287999735938178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,5120,0.011552000211344825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,2048,0.0076248885856734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,4096,0.011918221910794577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,2560,0.007979555262459649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,3072,0.016932444439993966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,2560,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,1536,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,2048,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,2048,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,2560,0.015224888920783997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,1024,0.004534222185611725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,1536,0.006156444549560547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,2048,0.014522666732470194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,768,0.003875555677546395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,1536,0.012727999852763282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,1024,0.0107360002067354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,768,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,512,0.003100444459252887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,768,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,512,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,256,0.0028364445186323593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,512,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,512,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,256,0.004973333328962326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,256,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,128,0.0025395556456512874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,4096,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,1536,0.006824888702895906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,64,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,128,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,256,0.009671111073758867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,32,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,128,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,1024,0.005650666852792104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,65536,0.12483378251393636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,128,0.007330666813585494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,64,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,32,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,768,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,51200,0.0983733336130778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,65536,0.07978044615851508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,16384,0.0380639996793535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,65536,0.06538399722841051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,51200,0.0525493323802948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,51200,0.06061600314246284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,12288,0.030718220604790583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,16384,0.020946666598320007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,16384,0.02456533412138621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,10240,0.02671555512481266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,12288,0.01847644481394026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,65536,0.1353706651263767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,51200,0.1084444456630283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,12288,0.017310221989949543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,8192,0.019575110740131803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,16384,0.04305244485537211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,7168,0.01642400026321411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,10240,0.015469332536061605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,10240,0.01647644407219357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,8192,0.013767110804716745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,6144,0.015298666225539314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,8192,0.014884443746672736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,7168,0.012735111018021902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,7168,0.013179555535316467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,10240,0.03068888849682278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,5120,0.013257777525318993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,8192,0.026371555195914373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,6144,0.011744889120260874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,6144,0.012130666938092975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,4096,0.011351111034552256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,7168,0.024263999528355066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,5120,0.01053155544731352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,3584,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,5120,0.010911111202504901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,6144,0.02255200015174018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,4096,0.009032888544930352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,3072,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,4096,0.010012444522645738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,5120,0.020446222689416673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,3584,0.008661333057615492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,2560,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,3584,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,4096,0.018235555953449674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,3072,0.008035555481910706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,2048,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,3584,0.017236444685194228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,3072,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,2560,0.007621333003044128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,2560,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,3072,0.016833777228991192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,12288,0.03534311056137085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,2048,0.007328888608349695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,2048,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,2560,0.01554666625128852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,1536,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,1536,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,2048,0.013067555924256643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,768,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,1024,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,1536,0.01274222218328052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,512,0.0028524444335036804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,768,0.005967111223273807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,1024,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,512,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,768,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,256,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,512,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,512,0.010339555641015371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,256,0.004984000076850255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,128,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,256,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,1024,0.0038382220599386427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,256,0.009397333694828881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,64,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,128,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,128,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,32,0.0021848889688650766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,1024,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,65536,0.11611821916368271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,128,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,64,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,32,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,768,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,51200,0.0895528859562344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,65536,0.0717502236366272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,65536,0.062498668829600014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,16384,0.03228800164328681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,51200,0.05510044429037306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,51200,0.05040533343950907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,12288,0.0266951107316547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,16384,0.020104888412687514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,16384,0.02216533323129018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,1536,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,51200,0.10752533541785346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,10240,0.02323377794689602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,16384,0.042675554752349854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,12288,0.01682577696111467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,8192,0.01922044489118788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,10240,0.015223999818166098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,10240,0.014875556031862894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,12288,0.03442755672666762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,8192,0.013223999904261695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,10240,0.030534221066368952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,8192,0.026161778304311965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,7168,0.012432888978057437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,6144,0.015503111812803479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,7168,0.0123848890264829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,7168,0.023883556326230366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,5120,0.01203644441233741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,6144,0.011349333657158745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,6144,0.011726222104496427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,12288,0.01701955497264862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,5120,0.009719111025333405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,6144,0.02208799951606327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,5120,0.010281778044170803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,4096,0.010953777366214328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,3584,0.009372444616423713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,4096,0.009126221968068017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,5120,0.020049777295854356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,4096,0.009487999810112847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,3072,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,7168,0.01752444439464145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,8192,0.014106666048367819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,3584,0.008240888516108194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,4096,0.01832177738348643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,2560,0.006907555378145642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,3072,0.007920000288221572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,3584,0.01717333330048455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,2048,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,2560,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,2560,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,3072,0.016191111670600045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,1536,0.0046657779150538975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,2048,0.006646222124497096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,65536,0.13525244924757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,2048,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,1024,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,1536,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,1536,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,2048,0.013430222041077085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,768,0.003192000091075897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,1024,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,1024,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,1536,0.011749332977665795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,512,0.003077333379122946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,3584,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,768,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,768,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,1024,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,256,0.002512000087234709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,512,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,512,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,768,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,128,0.002190222135848469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,256,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,256,0.004979555391603046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,64,0.0021671110557185277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,512,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,32,0.0021688888470331826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,128,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,256,0.009640889035330879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,128,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,2560,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,65536,0.08742133114073013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,128,0.007459555235173967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,64,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,32,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,51200,0.06993510988023546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,65536,0.06636266575919257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,65536,0.06050311194525825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,51200,0.05073600014050802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,51200,0.048341333866119385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,65536,0.13464889261457655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,12288,0.021458667185571458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,16384,0.019766221443812054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,16384,0.01957155598534478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,51200,0.10790133476257324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,10240,0.0177182224061754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,3072,0.007682667010360294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,12288,0.016146666473812528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,16384,0.0422017772992452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,12288,0.01584533353646596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,8192,0.01499111122555203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,10240,0.01460000044769711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,10240,0.014299554957283868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,12288,0.03411644366052415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,7168,0.013144888811641268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,8192,0.013187555803192986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,8192,0.012738666600651212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,6144,0.01167022188504537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,10240,0.03001333276430766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,8192,0.025901334153281316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,7168,0.011692444483439127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,5120,0.010250666903124915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,16384,0.027274666561020747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,6144,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,7168,0.024113777610990737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,4096,0.009453333086437648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,5120,0.009331555830107795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,6144,0.02232444451914893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,4096,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,4096,0.008301333420806462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,3584,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,5120,0.02017511096265581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,3584,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,4096,0.01790844400723775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,3072,0.007195555501514011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,3584,0.008063111040327284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,3584,0.017529777354664274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,3072,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,2560,0.006265777680608962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,3072,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,7168,0.011367110742463006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,2048,0.00546577779783143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,2560,0.007502222226725684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,2560,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,6144,0.010728889041476779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,1536,0.004418666578001446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,2560,0.014743111199802823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,2048,0.006226666685607698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,5120,0.00978577799267239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,1024,0.0031866667171319327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,1536,0.005967111223273807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,2048,0.013054221868515015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,768,0.0028408887899584244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,1024,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,1536,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,768,0.005395555661784277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,512,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,768,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,1024,0.010718222293588849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,768,0.010229333407349056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,512,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,256,0.002472888885272874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,512,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,3072,0.015093333191341825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,128,0.0021199999998013177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,512,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,256,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,256,0.005140444470776452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,64,0.0019128889673286015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,2048,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,32,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,128,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,256,0.009397333694828881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,1536,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,65536,0.0754515528678894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,64,0.005072000126043956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,128,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,32,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,51200,0.060421334372626416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,1024,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,16384,0.02436622149414486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,65536,0.06302311023076375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,65536,0.059398220645056836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,51200,0.04752622379197014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,51200,0.04870577653249105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,12288,0.01925511161486308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,16384,0.01925066610177358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,65536,0.13464978006150988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,10240,0.01649777756796943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,51200,0.10814666748046875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,12288,0.015253333581818474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,12288,0.015835555063353646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,12288,0.033512890338897705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,10240,0.01424088908566369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,8192,0.014089778065681458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,128,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,10240,0.029807110627492268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,8192,0.012273777690198688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,7168,0.012448888685968188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,8192,0.02602133320437537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,7168,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,6144,0.011053333679835001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,7168,0.011563555234008364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,7168,0.024078221784697637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,16384,0.018727999594476488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,5120,0.009648000200589498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,6144,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,6144,0.010543111297819348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,16384,0.041843556695514254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,4096,0.008947555389669206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,5120,0.009739555418491364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,5120,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,10240,0.01387555566098955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,3584,0.00758844448460473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,4096,0.008276444342401292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,4096,0.008231110870838165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,5120,0.020153777466879953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,3072,0.006895111252864202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,8192,0.012396444049146442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,3584,0.008211555580298105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,3584,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,4096,0.017602667212486267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,2560,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,3072,0.007668444679843054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,3072,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,2048,0.005125333451562458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,2560,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,3584,0.016185777054892648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,2560,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,1536,0.004446222136418025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,3072,0.014787554740905762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,2048,0.0069102222720781965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,1024,0.0033857776886887024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,2560,0.013803555733627744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,2048,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,1536,0.00675644435816341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,1536,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,768,0.002891555635465516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,2048,0.012768000364303589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,1024,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,512,0.002510222295920054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,1024,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,1536,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,6144,0.021706667211320665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,768,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,256,0.002190222135848469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,768,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,1024,0.011067555182509951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,512,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,128,0.0021377777059872947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,512,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,768,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,64,0.0018604445374674266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,256,0.004986666556861666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,32,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,128,0.005769777629110549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,512,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,65536,0.06696089108784993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,128,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,256,0.009011555876996782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,64,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,128,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,32,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,65536,0.05982310904396904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,65536,0.05855821900897556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,51200,0.04694933361477322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,51200,0.046655111842685275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,12288,0.016537777251667447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,65536,0.13464177979363337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,16384,0.018911111685964797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,16384,0.019172444939613342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,10240,0.014481777946154276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,51200,0.10777778095669216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,12288,0.015849777393870883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,16384,0.04138933287726508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,8192,0.012479999827014076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,10240,0.013450666434235044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,12288,0.03326844506793552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,8192,0.012402666939629449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,8192,0.011996444728639392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,10240,0.03009066647953457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,7168,0.011643555429246692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,51200,0.053103112512164645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,7168,0.010671111444632212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,6144,0.011632888681358762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,8192,0.025785777303907607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,16384,0.020177778270509508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,7168,0.011194667054547204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,5120,0.009640889035330879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,6144,0.009769777456919352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,7168,0.024119999673631456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,6144,0.010113777385817634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,4096,0.008440888590282863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,5120,0.009027555584907532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,5120,0.008979555633332994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,6144,0.02197866638501485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,3584,0.007240888973077138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,12288,0.015051555302408008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,4096,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,4096,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,5120,0.019868445065286424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,3072,0.006792888873153263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,10240,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,3584,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,3584,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,2560,0.00600622221827507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,4096,0.017508443858888414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,3072,0.007637333538797166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,3072,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,2048,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,3584,0.015923556354310777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,2560,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,1536,0.004197333421972063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,2560,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,3072,0.01477688882086012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,2048,0.00700711127784517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,1024,0.003522666792074839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,2048,0.006335999816656113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,2560,0.014152889450391134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,1536,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,768,0.003169777699642711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,1536,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,2048,0.01311377767059538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,1024,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,512,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,1024,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,1536,0.01202933324707879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,768,0.00601244428091579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,768,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,256,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,1024,0.010451555252075195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,512,0.005719111197524601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,512,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,128,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,256,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,256,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,64,0.002159111201763153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,256,0.009347555538018545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,32,0.0021893332401911416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,128,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,65536,0.055769774648878306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,128,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,64,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,32,0.0053635554181204895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,65536,0.057224889596303306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,65536,0.056735111607445605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,16384,0.020918221937285528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,51200,0.04489511251449585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,51200,0.04567111200756497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,65536,0.1344577736324734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,12288,0.015543111496501498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,16384,0.018791110979186166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,51200,0.10800000031789143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,768,0.0099973330895106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,10240,0.013752000199423896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,12288,0.01475911173555586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,12288,0.015501333607567681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,512,0.010048888623714447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,16384,0.0414373344845242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,8192,0.011730666789743634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,10240,0.013417777915795645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,10240,0.013286222186353473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,12288,0.033422221740086876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,7168,0.012313777373896705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,8192,0.012057777908113269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,8192,0.011377777490350934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,51200,0.046133332782321505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,6144,0.011352889239788055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,7168,0.010312889185216691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,8192,0.02572533322705163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,7168,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,5120,0.009197333620654212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,6144,0.009687999884287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,6144,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,4096,0.008176888856622908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,5120,0.00886222223440806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,5120,0.008421333299742805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,6144,0.02188355558448368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,3584,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,16384,0.017684444785118103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,5120,0.019485332899623446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,4096,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,4096,0.007670222057236566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,3072,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,3584,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,3584,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,2560,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,4096,0.017536888519922893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,3072,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,3072,0.007663999994595845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,3584,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,2048,0.00452533322903845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,3072,0.0148053334818946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,2560,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,2560,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,10240,0.029361777835422095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,1536,0.0041724443435668945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,2048,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,2048,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,2560,0.013748444616794586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,1024,0.0032764443506797156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,1536,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,1536,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,7168,0.02331377731429206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,768,0.003176888864901331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,2048,0.012785777449607849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,1024,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,1536,0.012063999970753988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,512,0.0028044444819291434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,1024,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,768,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,256,0.0024622221373849446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,768,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,1024,0.011010666688283285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,512,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,128,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,512,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,768,0.010058666268984476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,256,0.005519999812046687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,64,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,256,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,512,0.010033778018421596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,32,0.0021191111041439907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,128,0.005677333308590784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,128,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,65536,0.04615999923812019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,256,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,51200,0.03840444485346476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,64,0.004975111120276981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,128,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,32,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,65536,0.055600000752343066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,16384,0.015098666151364645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,51200,0.04468799961937798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,51200,0.04396888944837782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,12288,0.013754666679435305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,16384,0.017440888616773818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,65536,0.13452266322241888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,16384,0.01830400029818217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,10240,0.012381333443853589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,51200,0.10727555884255303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,16384,0.04120622078577677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,12288,0.01479288935661316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,12288,0.014383999837769402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,8192,0.011559111376603445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,10240,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,10240,0.012729778057999082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,7168,0.010318222145239512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,12288,0.03324000040690104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,8192,0.010958222051461538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,10240,0.029454221328099568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,6144,0.009344889058007134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,7168,0.010093332992659675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,7168,0.010144888526863521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,65536,0.05555999941296048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,6144,0.009073778159088558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,5120,0.008711111214425828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,5120,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,6144,0.021317332983016968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,4096,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,5120,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,5120,0.01958933307064904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,4096,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,3584,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,4096,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,3584,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,3072,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,8192,0.011243555280897351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,3584,0.016146666473812528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,8192,0.025348444779713947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,2560,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,3072,0.0069848886794514125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,3072,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,6144,0.009468444519572789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,3072,0.01443199978934394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,2048,0.004559999952713649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,2560,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,2048,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,1536,0.0038782221575578055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,2048,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,2048,0.013103110922707452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,1536,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,1024,0.0031848889258172777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,4096,0.016844444804721408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,3584,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,1536,0.011413333316644033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,768,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,1024,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,768,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,1024,0.010690666735172272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,7168,0.022898667388492163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,512,0.002542222125662698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,2560,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,768,0.006142222219043308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,2560,0.013456888496875763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,512,0.005986666513813867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,256,0.0022026666750510535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,512,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,768,0.010006222460005019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,128,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,256,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,512,0.00941244430012173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,256,0.004947555561860402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,64,0.0020071110791630214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,1536,0.006521777974234686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,128,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,32,0.0020151111400789684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,128,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,256,0.009339555270142024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,65536,0.036702222294277616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,64,0.005049777941571342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,1024,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,128,0.0075902218619982404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,51200,0.029504888587527808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,32,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,65536,0.05481600099139743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,65536,0.05307999915546841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,51200,0.04445777667893303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,51200,0.041927112473381885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,12288,0.012072888513406118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,16384,0.017631111873520743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,51200,0.10708177751964992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,16384,0.04062311185730828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,10240,0.010431110858917236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,12288,0.014138667119873894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,12288,0.013740444348918067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,8192,0.009668444593747457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,10240,0.012079111403889127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,12288,0.03292355603641934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,8192,0.01089777797460556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,8192,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,10240,0.029254221253924902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,7168,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,8192,0.025388444463411968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,7168,0.01037333326207267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,6144,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,7168,0.009455111291673448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,16384,0.016881777180565726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,6144,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,6144,0.008939555949634975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,6144,0.020997333857748244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,5120,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,4096,0.00721599989467197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,5120,0.008833777573373582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,5120,0.018572444717089336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,10240,0.012353777885437012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,3584,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,4096,0.0074746666683091064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,4096,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,3584,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,4096,0.01717688971095615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,16384,0.0138568886452251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,3584,0.015448000695970325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,3072,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,7168,0.022681777675946552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,2560,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,5120,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,3072,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,2048,0.004337777694066365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,3072,0.0144195556640625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,2560,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,2560,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,1536,0.003834666477309333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,2048,0.006821333534187741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,2048,0.006223111102978389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,2560,0.014135999812020196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,1024,0.003181333343187968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,1536,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,2048,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,1536,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,768,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,1024,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,1024,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,1536,0.012111999922328524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,3584,0.007108444141017065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,512,0.0024844445288181305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,768,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,1024,0.010473777850468954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,256,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,512,0.005668444352017508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,65536,0.1342924435933431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,768,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,128,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,256,0.004969777746333016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,256,0.00565244464410676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,128,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,128,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,64,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,256,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,32,0.0018853334089120228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,128,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,65536,0.0295724438296424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,64,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,32,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,3072,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,51200,0.02421155571937561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,768,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,65536,0.05434933304786682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,512,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,51200,0.04387022058169047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,65536,0.13383732901679143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,16384,0.01109688894616233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,51200,0.0407244430647956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,12288,0.009398221969604492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,512,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,16384,0.016196444630622864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,51200,0.10699555608961318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,16384,0.016173332929611206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,10240,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,12288,0.012906666431162091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,12288,0.013312000367376538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,16384,0.04047822289996677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,8192,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,10240,0.012067555553383298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,10240,0.011367999845080905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,12288,0.03273866573969523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,7168,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,8192,0.010334221853150262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,8192,0.010738666686746808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,10240,0.028794666131337483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,6144,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,7168,0.009673777553770278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,7168,0.00999644481473499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,65536,0.050632887416415744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,5120,0.006903999795516332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,6144,0.009002666506502364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,6144,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,7168,0.02232355541653103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,4096,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,5120,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,6144,0.020562666985723708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,3584,0.006207110981146495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,4096,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,4096,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,5120,0.018193778064515855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,3072,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,3584,0.007840889195601145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,3584,0.007032000356250339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,3072,0.0069120000633928515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,3072,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,2560,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,3584,0.01578044394652049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,2048,0.0042257776690853965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,2560,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,3072,0.015145777000321282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,2560,0.0064106666379504735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,8192,0.024745777249336243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,2048,0.006116444451941385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,2048,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,2560,0.013416888813177744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,1024,0.0031999999450312722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,1536,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,1536,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,2048,0.012774222426944308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,5120,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,768,0.0028808888875775864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,1024,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,1024,0.006060444646411472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,512,0.002472888885272874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,1536,0.01203733351495531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,768,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,768,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,4096,0.017160000072585214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,256,0.002130666747689247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,1024,0.011050666371981302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,768,0.010031110710567897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,128,0.0021093332519133887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,256,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,256,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,512,0.010037333601050908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,64,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,256,0.009308444129096137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,128,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,32,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,65536,0.023631110787391663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,128,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,64,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,32,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,51200,0.01941333380010393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,65536,0.0539991127120124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,51200,0.0399600002500746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,65536,0.13414755132463244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,16384,0.010049777726332346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,512,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,51200,0.04336000151104397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,51200,0.10633599758148193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,12288,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,16384,0.015174221661355762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,16384,0.015993777248594496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,128,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,1536,0.0038115556041399636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,10240,0.007696000238259633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,12288,0.012730666332774691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,16384,0.040087110466427274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,12288,0.013428444663683573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,8192,0.006656000183688269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,10240,0.011979555918110741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,12288,0.03256444467438592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,10240,0.011352000137170156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,65536,0.049639999866485596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,7168,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,8192,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,8192,0.01015822258260515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,10240,0.02812888887193468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,6144,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,7168,0.010033778018421596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,7168,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,8192,0.0243857784403695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,5120,0.006650666809744305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,512,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,6144,0.009376000199053023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,6144,0.008623110751310984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,4096,0.006055111272467508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,6144,0.0206986665725708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,5120,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,3584,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,4096,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,4096,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,5120,0.018527110417683918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,3072,0.005537777725193236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,4096,0.01648622254530589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,3584,0.007885333564546373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,3584,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,2560,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,3072,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,3584,0.01573155489232805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,2048,0.004252444538805219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,2560,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,3072,0.014829332629839579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,2560,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,1536,0.003847111016511917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,2048,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,2560,0.014113777213626437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,7168,0.02234755622016059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,1024,0.0031600000543726813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,1536,0.0058853332367208265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,1536,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,5120,0.0086986662613021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,768,0.0028248888750871024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,1024,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,1024,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,1536,0.011762667033407422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,512,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,768,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,768,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,1024,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,256,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,3072,0.006787555499209299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,512,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,768,0.010410666465759277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,128,0.0019271110908852683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,256,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,256,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,512,0.009670221971140968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,64,0.001812444378932317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,128,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,768,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,256,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,2048,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,65536,0.01699733402993944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,128,0.007346666521496243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,2048,0.01219555570019616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,51200,0.014568888478808932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,32,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,65536,0.05304444498485989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,16384,0.007664889097213745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,65536,0.049045334259668984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,51200,0.03873511155446371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,51200,0.03939022289382087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,65536,0.13382932874891493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,12288,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,16384,0.01499022212293413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,16384,0.015810666812790763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,10240,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,51200,0.1059066653251648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,512,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,12288,0.01311555587583118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,12288,0.012596444123321109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,8192,0.005773333211739858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,10240,0.012040889097584618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,10240,0.011428444749779172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,12288,0.03183822168244256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,7168,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,8192,0.010690666735172272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,8192,0.010017777482668558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,6144,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,10240,0.0277804434299469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,7168,0.010057777994208867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,8192,0.024328889118300542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,7168,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,5120,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,6144,0.008716444174448649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,7168,0.022628444764349196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,5120,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,5120,0.008807111117574904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,6144,0.020232000284724765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,5120,0.018906666172875296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,4096,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,16384,0.040087110466427274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,3584,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,4096,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,4096,0.016138666205936007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,3072,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,3584,0.007080888582600488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,3584,0.015799999237060547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,3072,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,2560,0.005207111024194294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,64,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,3072,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,6144,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,2048,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,3072,0.014470222923490735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,2560,0.007289778027269576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,4096,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,1536,0.003721777763631609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,2048,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,2048,0.006045333213276333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,1024,0.003185777821474605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,1536,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,2048,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,1536,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,768,0.0028808888875775864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,1024,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,1024,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,1536,0.011739555332395764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,3584,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,512,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,768,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,768,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,256,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,512,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,512,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,768,0.010336000058386061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,128,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,256,0.004949333353175057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,512,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,256,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,64,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,2560,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,512,32,0.0017742222795883815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,128,0.0053484443989064955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,65536,0.012532444463835822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,256,0.009373333719041612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,2560,0.01313511116637124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,64,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,51200,0.01110311100880305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,32,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,65536,0.04681333237224155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,65536,0.04650577902793884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,16384,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,51200,0.037428445286220975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,51200,0.036995556619432234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,65536,0.13324177265167236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,12288,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,1024,0.011027555498811932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,16384,0.01583377851380242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,51200,0.10549510849846734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,10240,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,12288,0.013298666311634911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,12288,0.012447111308574677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,16384,0.0392524434460534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,8192,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,10240,0.011067555182509951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,10240,0.01202666676706738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,12288,0.032187554571363665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,7168,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,8192,0.009744000103738572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,10240,0.028449777099821303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,128,0.007280888656775157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,6144,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,7168,0.009653333160612319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,8192,0.023992889457278784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,6144,0.009309333231714036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,7168,0.02199999988079071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,6144,0.008403555386596257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,5120,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,16384,0.01496088835928175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,4096,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,5120,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,6144,0.020633776982625324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,5120,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,3584,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,4096,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,5120,0.01825777855184343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,4096,0.007325333025720384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,3072,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,3584,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,4096,0.016174222032229107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,8192,0.010684444672531553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,3072,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,3584,0.015855999456511605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,2560,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,3072,0.0075191110372543335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,7168,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,2048,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,2560,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,3072,0.014471999473041959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,2560,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,1536,0.003519999898142285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,2048,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,2048,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,2560,0.013626666532622443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,1024,0.0028897778441508612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,1536,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,1536,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,2048,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,768,0.0026435556097163092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,1024,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,1024,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,1536,0.011422221859296164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,512,0.0024764444679021835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,3584,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,128,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,768,0.005978666659858491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,768,0.0053324442770746015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,256,0.0021333334346612296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,1024,0.011040888726711273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,512,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,512,0.005203555441564984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,128,0.0018275556051068837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,768,0.010223999619483948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,256,0.005087999833954705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,64,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,128,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,512,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,256,32,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,256,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,65536,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,128,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,32,0.004536889079544279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,51200,0.008376888930797577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,65536,0.045533332559797496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,65536,0.04597244328922696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,65536,0.13189866807725695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,51200,0.03734755516052246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,16384,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,51200,0.03611200054486593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,16384,0.015837333268589444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,51200,0.10526221990585327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,12288,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,256,0.00564533347884814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,16384,0.039432889885372586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,10240,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,12288,0.013154666456911298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,12288,0.012747555143303342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,8192,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,12288,0.03149866726663377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,10240,0.011772444678677453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,64,0.005075555708673266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,10240,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,7168,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,8192,0.01075644459989336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,10240,0.028338667419221666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,6144,0.005168000029193031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,7168,0.009389333426952362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,7168,0.010018666585286459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,8192,0.024059555596775476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,5120,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,6144,0.009379555781682333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,7168,0.022338666849666174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,6144,0.008829333715968663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,4096,0.004565333326657613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,5120,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,16384,0.0147724449634552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,5120,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,6144,0.020256888535287645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,3584,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,4096,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,4096,0.007347555624114142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,5120,0.01813333398765988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,3072,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,128,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,3584,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,2560,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,4096,0.016883555385801528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,3072,0.007346666521496243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,3584,0.015058666467666626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,2048,0.004544888933499654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,8192,0.010049777726332346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,2560,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,2560,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,3072,0.015145777000321282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,1536,0.003547555456558863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,2048,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,2560,0.013749333719412485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,2048,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,1024,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,1536,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,2048,0.013050666285885705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,1536,0.011765333513418833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,1024,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,768,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,1024,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,768,0.006083555519580841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,1024,0.010398222340477837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,768,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,3584,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,768,0.010392000277837118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,512,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,256,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,3072,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,512,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,128,0.001863111017478837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,256,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,512,0.009332444104883406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,256,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,64,0.0017884444031450485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,128,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,128,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,32,0.00179822225537565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,256,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,65536,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,1536,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,51200,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,16384,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,64,0.00461422238085005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,128,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,32,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,10240,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,65536,0.04519644379615784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,51200,0.03631822268168131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,8192,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,16384,0.014797333214018079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,7168,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,6144,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,128,512,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,12288,0.01236533373594284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,5120,0.004597333156400257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,10240,0.011025778121418424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,4096,0.00452533322903845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,3584,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,8192,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,7168,0.009307555854320526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,3072,0.004602666530344221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,2560,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,5120,0.007667555577225155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,2048,0.004416888786686791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,4096,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,1536,0.0035075553589397003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,3584,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,1024,0.0028764444092909494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,3072,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,768,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,2560,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,512,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,2048,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,256,0.0022026666750510535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,1536,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,128,0.0018586667461527719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,1024,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,12288,0.005192000004980299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,512,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,64,32,0.0017742222795883815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,256,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,65536,0.008951110972298516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,51200,0.007970666719807519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,128,0.004557333472702238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,64,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,16384,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,12288,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,10240,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,65536,0.04488977789878845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,51200,0.0362684428691864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,8192,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,7168,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,16384,0.014879110786649915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,6144,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,6144,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,10240,0.011030221978823343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,5120,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,4096,0.004253333227501975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,8192,0.009737778041097853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,7168,0.009024889104896123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,6144,0.008692444198661381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,3072,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,5120,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,4096,0.007505777809354994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,2560,0.004554666578769684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,2048,0.004208000169859992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,3584,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,1536,0.003273777870668305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,3072,0.006685333533419504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,1024,0.0028497777465316984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,2560,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,2048,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,1536,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,32,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,1024,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,768,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,256,0.0018071112119489247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,12288,0.012435555458068848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,128,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,512,0.005081777771313985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,64,0.0017884444031450485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,256,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,3584,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,32,0.0018088887963030075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,128,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,64,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,32,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,768,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,768,0.002551111082235972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,16384,0.36666401227315265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,64,32,512,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,16384,0.16648622353871664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,51200,0.49485688739352757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,12288,0.2779928843180339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,10240,0.23004444440205893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,12288,0.12375556098090278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,51200,1.1253848605685763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,12288,0.11623822318183051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,8192,0.18711200025346544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,7168,0.16536000039842394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,10240,0.10513599713643391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,51200,0.4200266732109918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,8192,0.07027555836571588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,10240,0.0919955571492513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,7168,0.07515822516547309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,8192,0.08888622124989827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,6144,0.13885154989030626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,5120,0.11795732710096572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,7168,0.06237688991758558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,6144,0.05454666746987236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,5120,0.04674488968319363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,6144,0.06628444459703234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,3584,0.08333510822719997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,5120,0.055528889099756874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,4096,0.03963111175431146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,3072,0.07378311289681329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,16384,0.16137777434455022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,4096,0.04622488882806566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,2560,0.06103822257783678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,3584,0.03564622335963779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,3584,0.04121866822242737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,3072,0.031200889084074233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,3072,0.036758220858044095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,1536,0.03895644346872965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,2560,0.027440889014138117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,1024,0.027661333481470745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,2560,0.03234133455488417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,2048,0.027493332823117573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,1536,0.018556444181336295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,768,0.02144444485505422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,1024,0.015000000596046448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,512,0.014448000325096978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,768,0.01350844403107961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,768,0.01366311146153344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,256,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,1024,0.016179554992251925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,512,0.010870222416188983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,4096,0.09475111299090916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,128,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,256,0.010496888723638324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,256,0.009000889129108852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,64,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,128,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,32,0.006339555399285422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,128,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,2048,0.049545778168572314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,64,0.009720889230569204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,32,0.012599999705950419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,2048,0.023185777995321486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,1536,0.022164444128672283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,16384,0.27512889438205296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,51200,0.3974924352433946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,512,0.01201600001917945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,65536,0.5246186786227757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,65536,1.0939564175075955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,51200,0.8516071107652453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,12288,0.20677867200639513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,16384,0.13366844918992785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,10240,0.17313688331180146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,12288,0.10014844603008694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,16384,0.12548266516791448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,51200,0.3329155445098877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,10240,0.07172889179653592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,8192,0.05676977501975166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,8192,0.07231555382410686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,8192,0.1404915518230862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,6144,0.10671466588973999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,7168,0.06142933501137627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,7168,0.05071466830041674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,5120,0.0899582240316603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,6144,0.05394133263164096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,4096,0.07247200277116564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,5120,0.04572355416085985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,5120,0.03763555486996969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,4096,0.038077334562937416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,4096,0.03199377655982971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,3584,0.06448888778686523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,12288,0.0829555524720086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,10240,0.08528177605734931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,3072,0.056286222404903836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,3584,0.034091555409961276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,2560,0.04733511143260532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,3072,0.026703110999531213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,3072,0.030179556873109605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,2048,0.038302222887674965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,2560,0.026464000344276428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,6144,0.04336888922585381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,2560,0.022875555687480505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,1536,0.030213332838482324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,2048,0.01889955500761668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,1024,0.022443556123309668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,2048,0.023000000251664057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,1536,0.016143111719025504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,768,0.016122667325867545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,1536,0.01810933318403032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,1024,0.013407111167907715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,512,0.011343110766675739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,1024,0.014019555515713163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,768,0.012390221986505719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,256,0.00740533322095871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,768,0.011682666838169098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,65536,0.4257333278656006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,7168,0.12386843893263076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,512,0.010431999961535135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,512,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,64,0.003879111260175705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,128,0.004535111288229625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,128,0.009688888986905416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,256,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,32,0.0042035554846127825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,128,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,64,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,32,0.010777778095669217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,65536,0.3396906587812636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,65536,0.17536355389489067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,51200,0.27342844009399414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,16384,0.08283022377226087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,51200,0.13454400168524847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,65536,0.15023022227817112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,12288,0.0628346668349372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,3584,0.02914311157332526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,51200,0.12112177742852105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,10240,0.05288088983959622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,12288,0.0322106679280599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,12288,0.037316444847318865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,256,0.009328888522254096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,16384,0.04212799999448988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,8192,0.044277333550983004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,7168,0.03795911206139459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,10240,0.029211554262373183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,6144,0.03328444560368856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,10240,0.03230488962597317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,8192,0.0240631103515625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,5120,0.02843377656406826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,7168,0.021535111798180476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,7168,0.02433066732353634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,6144,0.019704888264338177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,8192,0.02769777841038174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,5120,0.02020088831583659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,3584,0.020615110794703167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,6144,0.0218213333023919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,3072,0.018343110879262287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,4096,0.0159244438012441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,3584,0.013904000322024027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,2560,0.015429332852363586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,3072,0.013293333351612091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,4096,0.015659555792808533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,2048,0.012686221963829465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,2560,0.0120319997270902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,16384,0.04899377624193827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,2560,0.012888888518015543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,1536,0.010195555786291758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,2048,0.010957332948843638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,1024,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,1536,0.009162666896979014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,1536,0.010079999764760336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,768,0.006339555399285422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,1024,0.007900444169839224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,1024,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,512,0.0047475554876857335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,4096,0.02439555525779724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,768,0.007365333537260692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,768,0.006521777974234686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,256,0.0032186667538351486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,512,0.0069848886794514125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,128,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,256,0.0063866666621632045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,3584,0.01459200017982059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,64,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,128,0.00674044465025266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,32,0.0025457777082920074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,128,0.005382222019963794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,64,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,32,0.006893333461549547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,2048,0.011416888899273343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,65536,0.2702844407823351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,51200,0.21171466509501138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,65536,0.15240000353919134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,5120,0.018921777606010437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,65536,0.132313781314426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,16384,0.07532888650894165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,512,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,12288,0.05180889036920336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,256,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,51200,0.11640089088016087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,51200,0.10520978106392755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,16384,0.04256355431344774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,3072,0.013281777501106262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,10240,0.04233244392606947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,16384,0.037238223685158625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,8192,0.035572446054882474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,7168,0.030999110804663763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,10240,0.02564711040920681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,12288,0.0325164430671268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,6144,0.027062222361564636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,8192,0.02184177769554986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,10240,0.028391109572516546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,8192,0.02482044365670946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,5120,0.023344000180562336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,7168,0.019550222489568923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,4096,0.01976355579164293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,7168,0.021702221698231165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,6144,0.01791911158296797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,3584,0.01698844465944502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,5120,0.01606755620903439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,5120,0.016526222229003906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,4096,0.014527110589875115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,3072,0.014870221416155497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,4096,0.014488000008794995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,3584,0.013449778159459433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,2560,0.012542222109105853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,2048,0.01088177743885252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,3072,0.012415111064910889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,3072,0.012299555871221753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,2560,0.011309333145618439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,1536,0.008619555168681675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,2048,0.009847111172146266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,2560,0.011471110913488599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,1024,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,2048,0.01036088913679123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,768,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,1536,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,1536,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,512,0.004230222354332606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,1024,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,1024,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,256,0.0030462222380770575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,768,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,512,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,128,0.0025351111673646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,512,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,256,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,256,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,6144,0.019183999962276883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,128,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,32,0.0026071110947264563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,64,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,3584,0.013286222186353473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,32,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,65536,0.21730044153001574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,51200,0.17315377129448783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,65536,0.14437066184149847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,51200,0.11050844192504883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,65536,0.12419022454155816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,12288,0.029150221082899306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,16384,0.062231997648874916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,768,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,12288,0.04124622212515937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,10240,0.034084445900387235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,51200,0.1013431085480584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,16384,0.0354693333307902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,12288,0.028340445624457464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,16384,0.04036533170276218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,8192,0.029120888974931505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,64,0.0025111111915773815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,7168,0.024780443973011438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,10240,0.024342222346199885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,10240,0.027077333794699773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,8192,0.021034666233592566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,128,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,8192,0.02328444520632426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,7168,0.018936889039145578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,7168,0.0203813330994712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,6144,0.017550221747822232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,4096,0.01608888970481025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,6144,0.018152889278199937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,3584,0.013517333401574029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,4096,0.013728000223636627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,4096,0.01404088901148902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,3072,0.012083555261294046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,3584,0.012826667063766055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,3584,0.013120888835854001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,2560,0.01035199976629681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,3072,0.011838222543398539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,3072,0.012087110843923358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,2560,0.010421333213647207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,12288,0.03115644388728672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,2048,0.008954666554927826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,6144,0.021722666091389124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,1536,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,2048,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,2048,0.010192888478438059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,1024,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,1536,0.008211555580298105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,5120,0.015400000744395785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,768,0.0042568888101312844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,1536,0.008178667061858708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,5120,0.01588444411754608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,1024,0.007301333049933116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,512,0.003556444413132138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,1024,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,768,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,256,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,512,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,768,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,128,0.002524444419476721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,512,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,256,0.006334222025341458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,64,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,32,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,256,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,128,0.006507555643717448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,128,0.0052266667286554975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,64,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,2560,0.011008000208271874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,65536,0.1759706603156196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,32,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,16384,0.04873866505093045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,65536,0.09308799770143296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,5120,0.01867733399073283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,65536,0.07078578074773152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,12288,0.03624000151952108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,51200,0.07210044728385077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,51200,0.05693688657548693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,16384,0.023022222850057814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,10240,0.029690666331185236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,16384,0.028761777612898085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,8192,0.025024889243973628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,12288,0.019310222731696237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,12288,0.02188355558448368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,10240,0.016516443755891588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,10240,0.019290667441156175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,8192,0.01442311041884952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,6144,0.019147555033365887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,7168,0.013575110998418597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,5120,0.01633155511485206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,7168,0.014774221513006421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,6144,0.012548444171746572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,4096,0.014205333259370593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,6144,0.013429332938459186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,5120,0.011412444214026133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,5120,0.01203555530971951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,3584,0.012103999654452005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,4096,0.010668444136778513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,3072,0.010776888993051318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,3584,0.009319110876984065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,51200,0.14061510562896729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,2560,0.00943911075592041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,3072,0.00889955543809467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,2048,0.00795022232664956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,3072,0.009498666558000777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,2560,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,1536,0.006851555572615729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,2560,0.009023110899660323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,2048,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,1024,0.005067555440796746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,2048,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,7168,0.021976000732845728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,768,0.0038817777401871155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,1536,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,1536,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,8192,0.016659556163681876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,512,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,768,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,1024,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,768,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,512,0.005966222120655908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,128,0.0024702221983008915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,512,0.0052986666560173035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,256,0.005664000080691443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,64,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,256,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,32,0.002275555498070187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,128,0.005886222339338726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,4096,0.010811555716726514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,128,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,64,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,3584,0.01016444464524587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,32,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,65536,0.08589688936869304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,65536,0.06720711125267877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,16384,0.0454675555229187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,51200,0.06539022260242038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,51200,0.05448711249563429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,12288,0.03558844327926636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,16384,0.021785777476098802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,1024,0.0064586665895250105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,16384,0.02637333340115017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,10240,0.02759733299414317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,256,0.0025724443710512584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,12288,0.017902221944597032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,8192,0.023759111762046814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,10240,0.01583377851380242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,12288,0.02019377715057797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,7168,0.020488889680968392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,10240,0.017622222503026325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,6144,0.017850667238235474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,8192,0.014138667119873894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,7168,0.012823999755912356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,8192,0.015691555208630033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,5120,0.015210666590266757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,7168,0.013834666874673633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,4096,0.013312888642152151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,6144,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,65536,0.16074400477939182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,5120,0.011008000208271874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,3584,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,4096,0.009879999690585667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,3072,0.010301333334710863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,51200,0.12658045026991102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,4096,0.010433778166770935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,3584,0.008887999587588841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,2560,0.009014222357008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,3584,0.00980266680320104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,2048,0.0075937774446275495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,3072,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,2560,0.007575111256705389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,3072,0.009296889106432596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,1536,0.006484444356626934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,2560,0.008390222158696916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,2048,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,1024,0.004763555609517627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,1536,0.006664888726340399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,768,0.0035537779331207275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,1536,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,1024,0.005640888793600931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,512,0.0031848889258172777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,768,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,512,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,512,0.005305777821275923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,6144,0.012721777790122561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,5120,0.011526222030321756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,128,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,256,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,256,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,64,0.002222222172551685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,128,0.0058373332851462895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,32,0.0024986666523747975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,128,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,64,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,2048,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,65536,0.13935644096798366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,32,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,51200,0.11014222436481053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,16384,0.03812444541189406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,768,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,65536,0.06467199987835355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,65536,0.07816533247629802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,12288,0.031016889545652602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,16384,0.020949333906173706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,51200,0.05961244636111789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,256,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,10240,0.02681866619322035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,12288,0.01715022159947289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,16384,0.02405333353413476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,8192,0.02347022294998169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,12288,0.018384888768196106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,7168,0.020608888732062448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,10240,0.015264888604482016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,10240,0.01629688839117686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,8192,0.013786666923099093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,7168,0.012830221818553077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,8192,0.014200000299347771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,7168,0.01309866706530253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,6144,0.011687111523416309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,4096,0.012251555919647217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,6144,0.012087110843923358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,5120,0.01072355525361167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,5120,0.01089511149459415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,4096,0.009104889300134447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,1024,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,4096,0.010066666536860997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,3584,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,3584,0.009402666654851701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,2560,0.008400888906584846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,3072,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,3072,0.008617777791288164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,2048,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,2560,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,2560,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,1536,0.005518222020732031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,2048,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,2048,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,5120,0.014104000396198697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,1536,0.006911110960774952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,1024,0.0041493334703975255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,1536,0.006231110956933763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,768,0.003253333270549774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,1024,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,3584,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,1024,0.005689777847793367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,512,0.0028906667398081887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,768,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,768,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,3072,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,256,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,128,0.0022239999638663395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,512,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,256,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,512,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,64,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,128,0.005648889061477449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,51200,0.05288266804483202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,256,0.005094222310516569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,32,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,128,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,65536,0.11700177192687988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,51200,0.09301777680714925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,64,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,32,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,65536,0.062197334236568876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,16384,0.03379199902216593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,65536,0.07108000251981947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,6144,0.01827555563714769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,12288,0.026862222287389968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,10240,0.022005332840813532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,16384,0.020237333244747586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,51200,0.05048355460166931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,16384,0.021903999977641638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,51200,0.05395199855168661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,8192,0.018239999810854595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,12288,0.016809778081046212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,7168,0.016608888904253643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,10240,0.014878221684032016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,6144,0.015426667200194465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,10240,0.01517688896920946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,7168,0.012478222449620565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,8192,0.013604444762070974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,5120,0.013234666652149625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,7168,0.012119111087587146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,6144,0.011402666568756104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,4096,0.01127199994193183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,5120,0.010072000324726105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,3584,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,5120,0.01038755559258991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,4096,0.008602666358153025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,3072,0.008990222381220924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,4096,0.009484444227483537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,3584,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,2560,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,3584,0.009016888837019602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,3072,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,2048,0.006127111199829314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,3072,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,2560,0.007681777907742395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,1536,0.00462666650613149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,2560,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,2048,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,1024,0.003760888758632872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,12288,0.016700444949997794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,2048,0.006526222245560751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,1536,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,768,0.003216888962520493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,512,0.0028853333658642242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,1536,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,1024,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,8192,0.01313688854376475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,256,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,1024,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,128,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,512,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,768,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,64,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,256,0.005709333138333426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,32,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,512,0.0052933332820733385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,128,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,256,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,65536,0.09512533081902398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,128,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,51200,0.0725511113802592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,64,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,16384,0.027662222584088642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,65536,0.059735112720065646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,65536,0.06469777557584974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,51200,0.048283557097117104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,12288,0.021420444051424663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,51200,0.049646221929126315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,10240,0.018538667096032035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,16384,0.01955911186006334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,16384,0.01959733333852556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,8192,0.01641688909795549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,12288,0.016185777054892648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,10240,0.014477334088749357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,12288,0.015584889385435315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,7168,0.014311111635631986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,8192,0.012713777522246042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,10240,0.014174222946166992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,8192,0.012770666844315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,6144,0.011644444531864591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,5120,0.010279110736317104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,6144,0.010188444621033138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,6144,0.010783111055692038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,4096,0.009123555488056606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,5120,0.009007111191749573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,6144,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,3584,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,4096,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,4096,0.008756444685988957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,32,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,3072,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,3584,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,2560,0.0063271112740039825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,3584,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,3072,0.007711110843552484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,2048,0.005399999933110342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,3072,0.00700711127784517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,2560,0.007252444823582967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,1536,0.004300444490379757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,2560,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,2048,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,1024,0.0032835555159383347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,2048,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,1536,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,7168,0.011528000235557556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,1536,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,1024,0.006325333482689328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,7168,0.01165600038237042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,512,0.002687999978661537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,1024,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,768,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,256,0.0025066667132907445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,5120,0.009757333331637913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,512,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,128,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,512,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,64,0.0021537778278191886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,128,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,256,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,32,0.0021342221233579847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,128,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,65536,0.0776968863275316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,64,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,32,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,51200,0.062096887164645724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,65536,0.05887999799516466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,65536,0.06222844123840332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,16384,0.0243857784403695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,12288,0.019824888971116807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,51200,0.04788177874353197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,768,0.0028977776981062363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,16384,0.019313777486483257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,16384,0.018954666124449838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,10240,0.016925333274735343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,8192,0.014919999572965833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,12288,0.0159599996275372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,12288,0.015152888165579902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,7168,0.012645333177513547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,8192,0.012424888710180918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,8192,0.012423111332787408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,6144,0.011066666907734342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,7168,0.011121778024567498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,7168,0.011306666665607028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,5120,0.009688888986905416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,6144,0.01054933336046007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,5120,0.008991999758614434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,5120,0.009311999711725447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,4096,0.008987555901209513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,51200,0.047264890538321606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,3584,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,4096,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,3584,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,10240,0.014420444766680399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,3072,0.006911110960774952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,768,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,3584,0.007631111476156447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,2560,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,256,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,2048,0.005121777868933148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,3072,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,2560,0.007350222104125553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,3072,0.0070266665683852295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,1536,0.004023111114899318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,2560,0.006781333436568578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,6144,0.01000533335738712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,2048,0.006983111302057902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,1536,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,2048,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,768,0.0028328889360030494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,1536,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,1024,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,512,0.00254133323000537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,1024,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,768,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,256,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,512,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,4096,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,128,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,256,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,256,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,64,0.0018257778137922287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,128,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,128,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,10240,0.013744889034165276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,32,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,65536,0.06878044207890828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,64,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,32,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,51200,0.05492888887723287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,65536,0.059650666183895536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,1024,0.003216888962520493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,65536,0.05788177914089627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,16384,0.02091022166940901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,12288,0.017478222648302715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,51200,0.04634488953484429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,16384,0.01901955571439531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,51200,0.04621155394448174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,10240,0.01477866702609592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,768,0.005971555494599872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,16384,0.018182223041852314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,8192,0.01296533313062456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,12288,0.01574844453069899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,10240,0.013854222165213691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,512,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,8192,0.011663999822404651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,10240,0.013200889031092325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,6144,0.010750222537252637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,8192,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,5120,0.009591999981138442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,7168,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,6144,0.009705777797434065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,7168,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,4096,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,6144,0.010093332992659675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,5120,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,3584,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,5120,0.008993777963850234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,3072,0.0068408888247278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,4096,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,2560,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,4096,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,3072,0.007748444047239091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,3584,0.007246221933099959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,2048,0.004839111119508743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,3072,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,2560,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,1536,0.004181333300140169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,2560,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,2048,0.00701422244310379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,1024,0.0035128887328836652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,2048,0.006207110981146495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,1536,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,768,0.003142222141226133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,1536,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,1024,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,512,0.0028195555011431375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,1024,0.005672889037264719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,256,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,512,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,512,0.0053022222386466126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,128,0.0021475555582178966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,256,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,256,0.004971555537647671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,64,0.0022088889446523455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,128,0.005538666827811136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,32,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,128,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,3584,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,64,0.004967111266321606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,32,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,51200,0.04702577657169766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,65536,0.056735999054378934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,65536,0.05740977658165825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,16384,0.01924711134698656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,12288,0.014780445231331719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,51200,0.045696000258127846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,51200,0.04424800144301521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,768,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,12288,0.01585244470172458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,10240,0.013665777941544851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,768,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,16384,0.0188711103465822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,16384,0.017532444662517972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,12288,0.015120888749758402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,8192,0.01199288914601008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,7168,0.011142222417725457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,10240,0.013187555803192986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,8192,0.011052444577217102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,10240,0.013111111190583972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,6144,0.010232888989978367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,7168,0.010403555300500657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,8192,0.011640888949235281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,5120,0.00923822240697013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,7168,0.010680889089902243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,6144,0.009520889156394536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,65536,0.058133330610063344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,4096,0.008216000265545314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,6144,0.00963199966483646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,3584,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,5120,0.008349333372380998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,4096,0.008268444074524773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,3072,0.0063164445261160536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,4096,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,3584,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,3584,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,2560,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,3072,0.0076248885856734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,2048,0.0047031111187405055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,3072,0.006788444601827198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,7168,0.012095111111799875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,2560,0.0074346669846110875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,1536,0.004001777619123459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,2560,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,2048,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,1024,0.0033893332713180115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,1536,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,2048,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,768,0.0032071111102898917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,12288,0.014427555931939019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,1536,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,512,0.002797333316670524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,768,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,1024,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,256,0.00249422217408816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,512,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,768,0.005437333136796951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,128,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,512,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,256,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,128,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,5120,0.008790222307046255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,32,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,256,0.004952888935804367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,128,0.004959110998445087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,65536,0.05032444331381056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,51200,0.03953244288762411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,64,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,32,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,65536,0.05551377932230631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,16384,0.016152888536453247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,65536,0.05441955394215054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,51200,0.04472622275352478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,12288,0.013810666898886362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,16384,0.018214222457673814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,10240,0.012456888953844706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,16384,0.017158221867349412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,12288,0.014143110977278816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,8192,0.011014222270912595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,12288,0.014118222726715935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,1024,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,10240,0.012455110748608908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,7168,0.01036266651418474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,10240,0.01276622215906779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,8192,0.010755555497275459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,6144,0.009565333525339762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,8192,0.010779555473062726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,7168,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,5120,0.008716444174448649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,7168,0.009671111073758867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,64,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,4096,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,6144,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,5120,0.008920888933870528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,3584,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,4096,0.008292444050312042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,3072,0.0057884446448749965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,4096,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,3584,0.007868444754017724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,2560,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,3584,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,3072,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,2048,0.004520888957712385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,2560,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,3072,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,1536,0.0038515557017591265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,51200,0.04267022344801161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,2560,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,1024,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,1536,0.006698666761318843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,768,0.0030053332448005676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,1024,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,1536,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,512,0.002521777732504739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,1024,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,768,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,256,0.0021377777059872947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,768,0.005281777845488654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,512,0.0053022222386466126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,128,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,256,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,256,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,64,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,5120,0.008221333225568136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,128,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,32,0.0019013333237833446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,128,0.004949333353175057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,65536,0.03821333249409994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,64,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,32,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,2048,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,51200,0.030436442957984075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,65536,0.05195822318394979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,2048,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,16384,0.01347644461525811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,51200,0.0443520016140408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,12288,0.01204622205760744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,16384,0.017581333716710407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,16384,0.01682044400109185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,512,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,10240,0.010603555374675326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,12288,0.013304000099500021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,6144,0.009546666509575313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,8192,0.009539555344316695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,12288,0.013735999663670858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,10240,0.01205599970287747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,7168,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,10240,0.012070222033394707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,8192,0.01070755554570092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,6144,0.008449777960777283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,7168,0.01036622209681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,8192,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,5120,0.007780444290902879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,7168,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,6144,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,4096,0.007086222370465596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,6144,0.00889866633547677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,5120,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,3584,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,5120,0.007991111112965478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,4096,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,3072,0.005531555662552516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,3584,0.007807999849319458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,65536,0.0547653337319692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,2560,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,3072,0.007500444849332173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,3584,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,2048,0.004508444418509801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,3072,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,2560,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,1536,0.003842666745185852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,51200,0.04105688797103034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,1024,0.003140444349911478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,2560,0.006452444526884291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,1536,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,2048,0.006135111053784688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,768,0.002834666727317704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,512,0.0025066667132907445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,1024,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,768,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,256,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,768,0.005452444569932089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,512,0.005893333504597346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,128,0.0020995556066433587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,512,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,256,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,64,0.0017991111510329777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,256,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,32,0.0019120000716712738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,128,0.00565155554148886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,128,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,4096,0.00758755538198683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,64,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,51200,0.023696889479955036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,32,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,65536,0.054435554477903575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,16384,0.011023999916182624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,65536,0.0495671100086636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,51200,0.043749332427978516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,2048,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,51200,0.039517333110173546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,10240,0.008630221916569604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,16384,0.01607377827167511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,12288,0.013129777378506131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,1536,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,8192,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,12288,0.012755555411179861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,7168,0.007335999773608313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,10240,0.012088889049159156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,10240,0.011384000380833944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,8192,0.010687111152542962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,6144,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,8192,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,7168,0.00998755544424057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,5120,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,6144,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,4096,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,5120,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,65536,0.027791112661361694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,5120,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,3584,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,4096,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,4096,0.007335111498832703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,3072,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,3584,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,3584,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,2560,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,3072,0.00739822205570009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,16384,0.0158595558669832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,2048,0.004520888957712385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,2560,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,2560,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,1536,0.00388355553150177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,2048,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,2048,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,1024,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,1536,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,1536,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,7168,0.009487999810112847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,768,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,1024,0.005974222388532426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,1024,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,512,0.0024968888610601425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,768,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,256,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,512,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,768,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,128,0.0021520000365045336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,512,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,256,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,64,0.001957333336273829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,256,0.005046222358942032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,128,0.005631999837027655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,12288,0.009354666703277165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,3072,0.00666133314371109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,64,0.005103111267089844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,32,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,51200,0.01939377850956387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,65536,0.04856977860132853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,65536,0.05413777629534403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,16384,0.010062221851613786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,51200,0.04309689005215963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,51200,0.03923199905289544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,12288,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,16384,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,16384,0.015209777487648858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,10240,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,12288,0.013112000293201871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,12288,0.012435555458068848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,8192,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,10240,0.011864888999197217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,32,0.0019484443797005548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,8192,0.010382222632567087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,7168,0.006800000038411882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,6144,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,65536,0.023316444622145757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,8192,0.010015111002657149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,5120,0.007985778152942657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,7168,0.0099982221921285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,6144,0.009319110876984065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,7168,0.009335111412737105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,4096,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,5120,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,3584,0.006231110956933763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,5120,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,4096,0.007970666719807519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,3072,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,4096,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,3584,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,2560,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,3584,0.0070159998204973005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,6144,0.008655110994974772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,2048,0.0041999999019834726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,3072,0.007364444434642792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,3072,0.006607111129495833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,2560,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,1536,0.003532444437344869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,2048,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,1024,0.002995555599530538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,2048,0.006082666830884085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,1536,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,768,0.0028239999794297745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,10240,0.011402666568756104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,1536,0.005838222387764189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,1024,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,512,0.0024560000747442245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,128,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,1024,0.005535111245181825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,256,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,512,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,128,0.0018604445374674266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,512,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,6144,0.008711111214425828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,64,0.0018035556293196147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,768,32,0.0018115554832749898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,256,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,128,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,65536,0.017092444830470614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,128,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,51200,0.014482667048772177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,64,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,65536,0.05235200126965841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,65536,0.04792888959248861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,16384,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,51200,0.03866577810711331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,51200,0.03868888815244039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,12288,0.007238222493065729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,2560,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,16384,0.015798222687509324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,10240,0.00646666685740153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,12288,0.013193777865833707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,12288,0.01238844460911221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,10240,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,768,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,768,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,10240,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,7168,0.00796977761718962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,8192,0.010424888796276517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,6144,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,8192,0.009952888720565373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,7168,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,5120,0.0069146665434042616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,7168,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,6144,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,4096,0.00656444455186526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,6144,0.00849333322710461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,32,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,5120,0.008732444710201686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,3584,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,5120,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,3072,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,4096,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,3584,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,4096,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,2560,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,3072,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,3584,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,2048,0.004256000121434529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,3072,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,2560,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,16384,0.014954666296641031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,1536,0.003617777592606015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,1024,0.0031537777847713898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,2048,0.006747555401590135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,1536,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,2048,0.006083555519580841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,768,0.0028044444819291434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,512,0.002500444443689452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,1536,0.005628444254398346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,1024,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,256,0.0021377777059872947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,512,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,768,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,128,0.002058666613366869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,512,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,256,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,64,0.0018071112119489247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,256,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,128,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,32,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,65536,0.014131555954615274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,128,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,64,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,51200,0.012138666378127204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,65536,0.04654844601949056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,65536,0.046204444434907704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,16384,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,51200,0.037556442949506975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,2560,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,51200,0.036335110664367676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,12288,0.006996444529957241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,512,8192,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,16384,0.015827554795477126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,10240,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,1024,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,16384,0.014793778459231058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,8192,0.006909333169460297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,768,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,10240,0.011706666813956367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,7168,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,8192,0.010412444670995077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,6144,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,8192,0.009752000371615091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,7168,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,5120,0.006309333360857434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,6144,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,6144,0.008404444489214156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,4096,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,5120,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,32,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,5120,0.00795377790927887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,3584,0.005972444597217772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,4096,0.00811288919713762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,3072,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,4096,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,3584,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,2560,0.004854222138722737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,3584,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,3072,0.0073928890956772705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,2048,0.004210666649871402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,3072,0.00674222202764617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,12288,0.01312355531586541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,2560,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,1536,0.0035128887328836652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,2560,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,12288,0.01238844460911221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,1024,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,2048,0.006807111203670502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,10240,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,2048,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,768,0.002612444468670421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,1536,0.006401777681377199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,512,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,1024,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,768,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,1024,0.005508444375462002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,256,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,768,0.0052879999081293745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,512,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,128,0.001962666710217794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,512,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,256,0.005639111002286275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,64,0.0018213333355055915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,256,32,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,256,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,65536,0.012717333104875354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,128,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,64,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,32,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,51200,0.011008000208271874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,65536,0.04618577824698555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,16384,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,65536,0.04423822296990288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,51200,0.037323554356892906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,12288,0.006265777680608962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,51200,0.03564622335963779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,16384,0.015586665934986539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,10240,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,16384,0.014952888091405233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,7168,0.00999111102686988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,12288,0.01311377767059538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,8192,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,12288,0.012406222522258759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,1536,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,7168,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,10240,0.011015999648306104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,8192,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,6144,0.005886222339338726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,7168,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,8192,0.009798222117953831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,5120,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,7168,0.008969777988062965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,6144,0.009320889082219865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,4096,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,128,0.0053386667536364655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,3584,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,6144,0.008657777474986183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,5120,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,4096,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,3072,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,4096,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,3584,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,2560,0.005005333158704969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,3584,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,3072,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,2048,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,3072,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,2560,0.007143110864692264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,1536,0.0035537779331207275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,1024,0.002970666728085942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,2560,0.006463999963468975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,2048,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,768,0.0028071111689011254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,1536,0.006435555716355641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,2048,0.005994666781690385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,512,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,1536,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,768,0.005877333382765452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,1024,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,768,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,512,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,128,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,512,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,256,0.005543111099137201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,64,0.0018159999615616268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,256,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,128,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,32,0.0017973333597183228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,5120,0.0086986662613021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,65536,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,128,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,51200,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,16384,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,64,0.004583999928500918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,12288,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,32,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,10240,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,65536,0.044165333112080894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,51200,0.03558222121662564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,8192,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,16384,0.014519110321998596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,7168,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,12288,0.012098666694429187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,6144,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,8192,0.009704888694816166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,1024,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,5120,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,7168,0.00944977750380834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,4096,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,10240,0.011761777930789523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,128,256,0.002128888956374592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,3072,0.004753777964247597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,6144,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,2560,0.004625777817434735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,5120,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,4096,0.007348444726732042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,3584,0.007255111303594377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,1536,0.003519111209445529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,3072,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,1024,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,2560,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,768,0.002524444419476721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,2048,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,512,0.0024986666523747975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,1536,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,256,0.002048888968096839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,1024,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,128,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,768,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,64,0.001820444439848264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,512,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,32,0.0018311111877361934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,256,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,10240,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,65536,0.012148444851239523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,128,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,51200,0.010687111152542962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,64,0.004560889055331548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,16384,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,12288,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,3584,0.005145777844720417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,32,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,10240,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,8192,0.005199999858935674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,65536,0.04420266548792521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,51200,0.03565244542227851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,7168,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,16384,0.014496889379289416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,6144,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,64,2048,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,12288,0.012406222522258759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,5120,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,10240,0.011239999698268043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,4096,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,3584,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,8192,0.009650666680600908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,3072,0.004554666578769684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,2560,0.004961777892377642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,7168,0.009022222624884712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,6144,0.008366222182909647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,2048,0.0041982221106688184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,1536,0.003220444545149803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,4096,0.007371555599901412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,5120,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,1024,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,3584,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,768,0.002546666603949335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,512,0.002198222196764416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,2560,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,256,0.0019519999623298645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,2048,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,1536,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,128,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,1024,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,64,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,768,0.005399111244413588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,49,32,32,0.0015031110702289476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,512,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,256,0.005078222188684675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,128,0.004584889031118817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,64,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,32,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,16384,0.36811110708448624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,16384,0.16571644941965738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,51200,0.49524354934692383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,3072,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,12288,0.2735662195417616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,51200,1.127263069152832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,12288,0.12445244524214004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,16384,0.13378755251566568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,10240,0.22851644621955025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,12288,0.11557332674662273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,10240,0.1052488883336385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,51200,0.4291368855370416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,8192,0.18859288427564833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,8192,0.08869244654973348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,10240,0.08998844358656143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,8192,0.06950400273005168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,7168,0.16152089171939427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,16384,0.2784702248043484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,12288,0.21306578318277994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,6144,0.14078933662838405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,7168,0.07506489091449313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,10240,0.17894932958814833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,7168,0.061443554030524365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,8192,0.14510044786665174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,5120,0.11871555116441514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,6144,0.06600444184409247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,6144,0.05412711037529839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,4096,0.09410133626725938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,5120,0.05539466606246101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,7168,0.12903111510806614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,5120,0.04672355453173319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,3584,0.0841440028614468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,6144,0.10926933421028985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,4096,0.04708266589376661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,4096,0.03889866669972738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,3072,0.07306933403015137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,3584,0.040185779333114624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,4096,0.07846489217546251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,3584,0.04126400086614821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,2560,0.06059200233883328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,5120,0.09425955348544651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,3072,0.03628355595800612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,3584,0.07051377826266818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,3072,0.031579554080963135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,2048,0.05029777685801188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,2560,0.026722666290071275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,2560,0.03205422229237027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,3072,0.06384711133109199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,1536,0.03973066806793213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,2048,0.022791999909612868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,2560,0.05961244636111789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,1024,0.02836266822285122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,2048,0.05046311020851135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,1536,0.022302221920755174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,1536,0.01870577699608273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,768,0.02102577851878272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,1024,0.01664799948533376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,1024,0.014907555447684394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,1536,0.0417484442392985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,768,0.013576889203654395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,768,0.013330666555298699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,1024,0.03366044494840834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,256,0.008260444634490544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,768,0.028571556011835735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,512,0.011141333315107556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,512,0.012177777787049612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,128,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,2048,0.027218666341569688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,512,0.024614221519894067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,256,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,64,0.00534400012758043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,128,0.00830666638082928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,128,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,256,0.020267556111017864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,32,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,64,0.009686222506894005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,128,0.015460444821251763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,51200,0.8552471266852485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,32,0.012709333664841123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,512,0.014582221706708273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,65536,0.5237778027852377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,51200,0.3971359994676378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,256,0.010464888479974536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,65536,1.0971511205037434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,51200,0.8647493256462945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,16384,0.2806008921729194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,16384,0.1337262259589301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,12288,0.21072977119021943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,65536,0.4157120121849908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,12288,0.10019377867380778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,16384,0.11515200138092041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,51200,0.32415734397040474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,12288,0.08665600087907578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,10240,0.17622933122846815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,10240,0.08507377571529812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,8192,0.14183378219604492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,10240,0.07134666707780626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,8192,0.055777774916754834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,8192,0.07156444258160062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,16384,0.22569423251681855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,7168,0.1254711151123047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,10240,0.1441377798716227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,7168,0.061428447564442955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,6144,0.10694399807188247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,7168,0.04908444484074911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,8192,0.11929778258005778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,5120,0.09095111158159043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,6144,0.043402665191226535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,7168,0.10533510976367527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,6144,0.05397955576578776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,5120,0.045796444018681846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,6144,0.09282400210698445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,5120,0.038471112648646034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,5120,0.07907999886406793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,12288,0.1735751099056668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,4096,0.03803999887572394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,4096,0.03265955713060167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,3584,0.06505955590142144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,3072,0.0565262238184611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,3584,0.03420266509056091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,3584,0.03422666589419047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,4096,0.0682328873210483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,3072,0.030199110507965088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,3584,0.06084799766540527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,3072,0.025943999489148457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,2560,0.04799555407630073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,4096,0.07411466704474555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,2560,0.026068444053332012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,2560,0.02676711148685879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,2048,0.03864622116088867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,3072,0.0559155543645223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,2048,0.01890222231547038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,1536,0.030543999539481267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,2048,0.023052444060643513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,1536,0.01603911154799991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,1536,0.01827911039193471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,1024,0.021910222040282354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,2048,0.042619556188583374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,1536,0.03601600064171685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,768,0.015923556354310777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,1024,0.013273778061072031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,1024,0.014144889182514615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,65536,0.9010302225748698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,768,0.01182577759027481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,1024,0.02925866511132982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,768,0.011922666596041786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,512,0.012091555529170565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,51200,0.7118435435824924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,256,0.00793422261873881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,512,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,128,0.0041724443435668945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,256,0.009405333134863112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,2560,0.04888622297181023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,256,0.007886221839321984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,64,0.0032142222755485107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,128,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,128,0.009727111293209923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,256,0.018518222702874076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,32,0.0035031110876136353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,128,0.0140328887436125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,64,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,32,0.010753778119881948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,65536,0.3491448826260037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,65536,0.18148977226681182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,51200,0.2715288798014323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,768,0.025060445070266724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,65536,0.15369155671861437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,512,0.010024000373151567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,16384,0.08162399795320299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,51200,0.1343582206302219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,512,0.021353777911927965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,51200,0.11627288659413655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,12288,0.06315200196372138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,16384,0.041879110866122775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,10240,0.052419556511773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,12288,0.035183111826578774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,12288,0.03749422232309977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,16384,0.05676355626848009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,8192,0.0439591109752655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,10240,0.03224533465173509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,10240,0.03274577856063843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,12288,0.04428977767626444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,7168,0.037738667594061956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,8192,0.02847466535038418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,8192,0.02851466668976678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,51200,0.14165511396196154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,65536,0.18920889165666369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,7168,0.024307555622524683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,6144,0.032928887340757586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,7168,0.021706667211320665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,8192,0.03473155697186788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,5120,0.028566221396128338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,6144,0.019855111837387085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,7168,0.0311173333062066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,6144,0.021895110607147217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,16384,0.04849777619043986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,5120,0.018816888332366943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,5120,0.01738755570517646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,3584,0.020411555965741474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,5120,0.025095999240875244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,4096,0.015440000428093804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,4096,0.016167999969588388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,3072,0.018210666047202218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,3584,0.015671999918089975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,4096,0.02188355558448368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,10240,0.03840444485346476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,2560,0.015275556180212231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,3584,0.02079288827048408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,3072,0.013337777720557319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,3072,0.01310222182008955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,2048,0.012719111310111152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,3072,0.018971555762820773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,2560,0.013976888524161445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,2560,0.01238933371173011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,1536,0.009970666633711921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,2048,0.010944888823562197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,2048,0.011704000333944956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,2560,0.01735822194152408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,1024,0.008051555189821456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,1536,0.009034666750166152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,6144,0.028200000524520874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,2048,0.016567111015319824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,1024,0.007956444389290279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,1536,0.014532445205582513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,768,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,1024,0.008066666622956594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,1024,0.01310222182008955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,768,0.007452444069915348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,512,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,768,0.007093333535724216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,512,0.006893333461549547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,768,0.011471110913488599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,256,0.002834666727317704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,512,0.005986666513813867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,512,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,256,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,128,0.002554666664865282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,256,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,4096,0.02352622151374817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,64,0.00222044438123703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,1536,0.009948444035318162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,256,0.009322666459613377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,32,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,128,0.005372444374693765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,64,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,128,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,3584,0.014763555592960782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,65536,0.2595253255632189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,65536,0.15410755740271673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,51200,0.2043057812584771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,65536,0.13268266783820257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,51200,0.11585066053602432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,16384,0.07321155733532377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,51200,0.10047199991014268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,16384,0.04217955470085144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,128,0.006814222368929122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,16384,0.036907556984159685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,16384,0.04786933461825053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,32,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,12288,0.03236177894804213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,65536,0.14678666326734754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,12288,0.03080444534619649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,10240,0.04245955414242215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,51200,0.11651466952429877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,12288,0.0376835564772288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,10240,0.025379555092917547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,10240,0.02830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,7168,0.031016889545652602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,8192,0.024677332904603746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,8192,0.024513777759340074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,6144,0.02701866626739502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,10240,0.03350044290224711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,7168,0.021669333179791767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,8192,0.030489779180950586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,7168,0.021939555803934734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,5120,0.023078221413824294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,6144,0.018050667312410142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,7168,0.02713511056370205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,12288,0.05201955636342367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,6144,0.0192577772670322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,4096,0.019675556156370375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,5120,0.016199999385409884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,5120,0.016506666938463848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,3584,0.01703555550840166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,6144,0.025095999240875244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,4096,0.01406222250726488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,4096,0.014377777775128683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,3072,0.014921777778201632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,5120,0.021687111920780603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,3584,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,3584,0.013434666726324292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,8192,0.03616177704599168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,2560,0.013088888592190213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,3072,0.01239377756913503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,2048,0.011112000379297467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,3072,0.012206222448084088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,2560,0.012265777422322167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,3072,0.017857778403494094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,2560,0.011121778024567498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,1536,0.008984888593355814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,2048,0.010575999816258749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,2560,0.016197333733240765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,1024,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,2048,0.015150222513410779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,1536,0.008352888955010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,1536,0.014740443891949125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,1024,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,1024,0.007328888608349695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,768,0.0058648888435628675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,1024,0.011401777466138205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,768,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,512,0.0041884444653987885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,768,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,4096,0.0199937770764033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,256,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,768,0.010675555302037133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,512,0.006769777586062749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,512,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,128,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,3584,0.01869777838389079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,256,0.006308444258239534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,512,0.009649777577983009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,256,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,2048,0.009830222361617619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,64,0.002483555633160803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,32,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,128,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,256,0.009387555221716562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,1536,0.009024889104896123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,64,0.005976888868543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,128,0.007264888948864407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,32,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,65536,0.21726844045850965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,65536,0.1451751126183404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,51200,0.17025689284006754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,16384,0.061102224720848926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,51200,0.11047022210227118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,51200,0.0966799987687005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,16384,0.040489776266945735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,16384,0.03728444377581278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,16384,0.04404977626270718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,128,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,51200,0.10838311248355442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,10240,0.033992889854643084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,12288,0.0310444434483846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,12288,0.030180444320042927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,10240,0.024515555964575872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,10240,0.027243556247817144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,12288,0.03597866826587253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,8192,0.028825776444541082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,65536,0.12031377686394586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,10240,0.03205688794453939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,7168,0.0255404445860121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,8192,0.021035555336210463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,8192,0.023375110493765935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,6144,0.022408889399634466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,7168,0.019271110494931538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,7168,0.020271110865804885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,8192,0.028631998433007136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,5120,0.01884711119863722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,12288,0.04201511210865445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,6144,0.01811377704143524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,6144,0.017204445269372728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,4096,0.015846222639083862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,7168,0.026313778426912095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,5120,0.015458666616015963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,5120,0.01588444411754608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,65536,0.13653511471218532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,3584,0.013760888742076026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,6144,0.024030221833123103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,4096,0.01499022212293413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,5120,0.021575111481878493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,3072,0.012479110724396177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,3584,0.013861333330472311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,2560,0.012264000044928657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,3584,0.013007111019558378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,3072,0.01202755586968528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,3072,0.012160888976520963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,3584,0.018175111876593698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,2048,0.01052177780204349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,2560,0.010739555789364709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,2560,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,1536,0.008623110751310984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,3072,0.01683555543422699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,2048,0.009029332962301042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,2048,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,2560,0.016168889072206285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,1024,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,1536,0.008055999875068665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,1536,0.008296888735559251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,2048,0.015543111496501498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,768,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,1024,0.007254222200976477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,1536,0.013776000175211163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,512,0.004545777622196409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,768,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,768,0.006595555692911148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,1024,0.011749332977665795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,4096,0.014081777797804939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,256,0.003160888950030009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,768,0.010672888822025724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,512,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,512,0.005656888915432825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,128,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,4096,0.01924977699915568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,256,0.006413333531883027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,64,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,128,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,512,0.00964355551534229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,32,0.002534222271707323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,128,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,256,0.009336888790130615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,128,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,64,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,32,0.0061928890645504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,51200,0.13779021633995905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,65536,0.09667022360695733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,65536,0.07017333639992608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,51200,0.07204711437225342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,16384,0.04724355538686117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,51200,0.05683022075229221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,16384,0.02795999911096361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,65536,0.1342408921983507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,256,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,12288,0.036031110419167414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,16384,0.02276266614596049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,51200,0.10811644130282932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,16384,0.04331466555595398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,10240,0.02979288829697503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,12288,0.01872177753183577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,12288,0.021920000513394673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,65536,0.17888177765740287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,8192,0.02544533378548092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,10240,0.016492444607946608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,10240,0.0192239996459749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,12288,0.035331557194391884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,7168,0.02197333342499203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,1024,0.006913777854707506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,8192,0.014588443769348992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,10240,0.031898667414983116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,6144,0.019107555349667866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,8192,0.027667555544111464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,7168,0.014452444182501899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,7168,0.01479200025399526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,5120,0.016187555260128446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,6144,0.013408000270525614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,7168,0.025254221426116094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,4096,0.013755555782053204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,5120,0.01202755586968528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,5120,0.011352000137170156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,6144,0.023279110590616863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,3584,0.012073777616024017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,4096,0.010503110786279043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,5120,0.02075111038155026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,3584,0.010113777385817634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,3584,0.010604444477293225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,3072,0.010980444649855295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,4096,0.018868444694413077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,3072,0.008951110972298516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,3584,0.017903111047214933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,3072,0.009476444787449306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,8192,0.016864889197879367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,3072,0.01719199948840671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,2560,0.00868622213602066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,2048,0.008012444608741337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,2560,0.009140444298585257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,6144,0.01241422196229299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,1536,0.006861333217885759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,2048,0.007631111476156447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,2560,0.015835555063353646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,2048,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,1024,0.004979555391603046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,1536,0.0069066666894488865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,1536,0.00666933341158761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,2048,0.014859555496109856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,768,0.0038764443662431505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,4096,0.011022222538789114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,1024,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,1536,0.013402666482660504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,1024,0.006001777946949005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,512,0.003157333367400699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,768,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,256,0.002542222125662698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,1024,0.010718222293588849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,512,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,768,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,128,0.002303999952144093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,2560,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,256,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,512,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,256,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,64,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,32,0.002206222257680363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,256,0.009300444689061906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,128,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,64,0.005514666852023866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,65536,0.15521866745418972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,32,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,128,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,65536,0.08621599939134385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,51200,0.06541866726345487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,65536,0.06705333126915826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,16384,0.04529066549407112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,768,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,512,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,51200,0.05428711242145962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,16384,0.02186311119132572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,12288,0.0354968872335222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,16384,0.02608000073168013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,65536,0.1344239976671007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,10240,0.027639110883076985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,51200,0.10731199714872573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,128,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,12288,0.017850667238235474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,12288,0.02013422217633989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,8192,0.023640000157886084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,10240,0.015847111741701763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,10240,0.01758577757411533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,12288,0.035199112362331815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,7168,0.020582222276263766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,8192,0.015395555231306287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,8192,0.014104000396198697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,51200,0.12267643875545925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,10240,0.03126488791571723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,6144,0.017749332719379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,8192,0.026713778575261433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,7168,0.013067555924256643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,7168,0.013810666898886362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,5120,0.015416888727082146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,6144,0.01278133359220293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,6144,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,7168,0.024450666374630396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,5120,0.011066666907734342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,5120,0.011420444481902652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,6144,0.02295822236273024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,3584,0.01128266668981976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,4096,0.009671111073758867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,5120,0.019996444384257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,4096,0.010240889257854885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,3072,0.010327999790509542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,16384,0.043037331766552396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,3584,0.009974222216341231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,4096,0.018585777944988675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,3072,0.008005333443482717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,3072,0.009208888643317753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,3584,0.01779733267095354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,3072,0.016531555189026725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,2560,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,2048,0.007676444119877285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,2560,0.00849777791235182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,2560,0.015215999550289579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,2048,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,1536,0.006200888918505774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,2048,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,2048,0.014527110589875115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,1536,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,1024,0.004620444443490771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,1536,0.01204711116022534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,3584,0.00903200027015474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,1024,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,768,0.0037911112109820047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,1024,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,2560,0.009005332986513773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,1024,0.011054221954610614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,512,0.0032142222755485107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,768,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,768,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,256,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,512,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,512,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,768,0.010052444206343757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,128,0.002293333411216736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,512,0.009704888694816166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,256,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,256,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,4096,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,32,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,128,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,256,0.008974221845467886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,1536,0.00639111093348927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,64,0.005399111244413588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,65536,0.13515378369225398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,32,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,128,0.007351111206743453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,51200,0.10668088992436726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,65536,0.0801102254125807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,65536,0.06504088640213013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,16384,0.03724977705213759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,51200,0.05927822324964735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,51200,0.05241688754823473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,12288,0.030046221282747056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,16384,0.02370577719476488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,65536,0.13409422503577337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,51200,0.10709333419799805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,16384,0.0422017772992452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,10240,0.02548622257179684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,12288,0.01851377718978458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,64,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,12288,0.017242666747834947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,8192,0.020826665891541373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,128,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,10240,0.015236443943447538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,12288,0.03456177645259433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,10240,0.016217778126398723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,7168,0.01846844454606374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,8192,0.0136835558546914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,6144,0.018107554978794523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,7168,0.01273333364062839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,10240,0.030236442883809406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,7168,0.01312800000111262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,8192,0.02630399995379978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,5120,0.012712000144852532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,6144,0.011753777662913004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,7168,0.024313777685165405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,6144,0.012048888537618848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,4096,0.011050666371981302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,5120,0.010614222122563256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,5120,0.010784000158309937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,3584,0.009745777481132084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,6144,0.022319111559126113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,4096,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,4096,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,5120,0.02028888960679372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,3072,0.008481777376598781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,3584,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,4096,0.018351111147138808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,3072,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,3072,0.00851733320289188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,3584,0.017196445001496207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,2560,0.007753777835104201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,8192,0.014463111758232117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,2048,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,2560,0.007452444069915348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,3072,0.01608888970481025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,2560,0.008278222547637092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,1536,0.005688000056478713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,2048,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,2560,0.015071110592948066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,2048,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,1024,0.0038773334688610504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,1536,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,1536,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,16384,0.021220443977249995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,768,0.0034808889031410217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,1024,0.005635555419656966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,1024,0.006503110958470239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,1536,0.01199288914601008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,512,0.0028159999185138275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,768,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,3584,0.00960177762640847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,768,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,1024,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,256,0.002461333241727617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,512,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,512,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,768,0.010715555813577441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,128,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,256,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,256,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,512,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,128,0.005624889085690181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,128,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,65536,0.11422399679819743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,128,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,64,0.0052879999081293745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,32,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,2048,0.013418667018413544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,51200,0.09059377511342366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,65536,0.06219644678963555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,65536,0.07086488935682508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,16384,0.033441778686311506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,51200,0.054040888945261635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,51200,0.05033066537645128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,65536,0.13355822033352321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,16384,0.021895999709765118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,12288,0.025776000486479864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,16384,0.020216888851589628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,51200,0.10695199833975898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,64,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,10240,0.02195466723706987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,16384,0.04200444287723965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,32,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,256,0.009300444689061906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,12288,0.016912889149453905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,8192,0.018066666192478605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,10240,0.014831110835075378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,7168,0.016579555140601266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,10240,0.015090665883488126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,8192,0.013334222137928009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,12288,0.03393244412210252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,8192,0.013801777528391944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,10240,0.030287110143237646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,7168,0.012345777617560493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,7168,0.012254222399658628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,5120,0.013081777426931592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,8192,0.026457778281635706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,6144,0.01130488928821352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,6144,0.011556444068749746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,7168,0.02397866712676154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,4096,0.011501332951916588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,5120,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,5120,0.010334221853150262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,6144,0.022237333986494277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,3584,0.00999644481473499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,4096,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,4096,0.009687999884287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,5120,0.02028888960679372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,3072,0.009229333036475712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,3584,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,3584,0.009026666482289633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,4096,0.018259555101394653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,12288,0.016674665941132438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,2560,0.007845333053006066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,3072,0.008375999828179678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,3072,0.007680888805124495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,2048,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,2560,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,2560,0.007187555233637492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,3072,0.016163556112183463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,1536,0.004857777721352047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,2048,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,2048,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,2560,0.014099554883109199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,1024,0.0037226668662495087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,2048,0.013775111072593264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,1536,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,1536,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,768,0.0032346666687064697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,1024,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,1024,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,1536,0.012115555504957834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,512,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,768,0.006109333286682765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,768,0.005541333307822545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,1024,0.011086222198274402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,256,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,512,0.0058311112225055695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,768,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,3584,0.016890666551060148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,128,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,256,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,256,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,512,0.009682666924264696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,64,0.0021084443562560612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,6144,0.01475911173555586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,128,0.005753777921199799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,32,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,256,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,128,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,65536,0.09102844529681736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,128,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,32,0.005406222409672207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,51200,0.07156978050867717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,65536,0.06471111377080281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,65536,0.06009511152903239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,16384,0.02753422160943349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,51200,0.04979911115434435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,51200,0.04843822121620178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,65536,0.13282844755384657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,12288,0.02128177715672387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,512,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,16384,0.019474666979577806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,16384,0.01954844428433312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,51200,0.10709866550233628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,10240,0.018582221534517076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,12288,0.016125332978036668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,12288,0.015684444043371413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,16384,0.04173688756095039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,10240,0.014107555150985718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,10240,0.014418666561444601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,12288,0.03349511159790887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,7168,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,64,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,8192,0.01274400038851632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,8192,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,10240,0.029454221328099568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,6144,0.01258666647805108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,7168,0.01165511127975252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,7168,0.011485333244005838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,5120,0.01013777818944719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,8192,0.026216889421145122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,6144,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,6144,0.010420444111029306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,7168,0.023727110690540735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,4096,0.009171555439631144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,5120,0.009646221995353699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,6144,0.021758221917682227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,5120,0.009761778016885122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,3584,0.007920888562997183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,4096,0.009146666361225976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,5120,0.02034133341577318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,3072,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,3584,0.00796622203456031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,3584,0.007670222057236566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,2560,0.0063279999627007385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,3072,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,3072,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,3584,0.016543999314308167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,2048,0.0053324442770746015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,2560,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,3072,0.015133332875039844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,2560,0.007242666350470648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,1536,0.004502222355869082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,2048,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,2048,0.006537777682145436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,2560,0.014383110735151501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,1024,0.0032631111227803757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,1536,0.006908444480763541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,1536,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,2048,0.013415999710559845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,768,0.0028897778441508612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,1024,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,1536,0.011735111474990845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,4096,0.008348444269763099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,768,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,512,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,1024,0.01072533345884747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,4096,0.018030222919252183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,256,0.0024844445288181305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,512,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,512,0.005290666802061929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,512,0.009657777845859528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,256,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,128,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,256,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,256,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,128,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,128,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,64,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,32,0.0022115556316243275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,1024,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,128,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,65536,0.07786489195293851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,768,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,32,0.005283555636803309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,8192,0.016135111451148987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,51200,0.06236977709664238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,65536,0.06107822391721937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,65536,0.05865866608089871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,16384,0.02325155503220028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,51200,0.04782488942146301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,51200,0.047424889273113675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,65536,0.1335902214050293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,12288,0.018984888990720112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,16384,0.019274666905403137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,51200,0.10654399792353313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,16384,0.04143822193145752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,12288,0.016064888901180692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,10240,0.016194666425387066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,12288,0.03304355674319797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,64,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,10240,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,8192,0.014083556003040738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,10240,0.014244443840450711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,768,0.010048888623714447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,7168,0.012784000072214337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,8192,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,8192,0.01238222254647149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,6144,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,8192,0.02587466604179806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,7168,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,7168,0.011369778050316704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,5120,0.009679111341635386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,16384,0.018927999668651156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,6144,0.01032977799574534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,6144,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,7168,0.023549333214759827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,4096,0.00886933339966668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,5120,0.009027555584907532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,12288,0.015173332558737861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,5120,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,3584,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,6144,0.0219217770629459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,4096,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,3072,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,3584,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,4096,0.017940445078743827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,3584,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,2560,0.006160000132189856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,3584,0.016879111528396606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,10240,0.029341333442264136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,3072,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,2048,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,2560,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,3072,0.015225778023401896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,2048,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,2048,0.006992888947327931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,1536,0.004431999805900785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,2560,0.014112888111008538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,1536,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,2048,0.0127368892232577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,1024,0.003216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,1536,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,4096,0.008758222063382467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,1536,0.011750222080283694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,5120,0.019827556278970506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,1024,0.011003555523024665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,768,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,3072,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,512,0.002776888923512565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,768,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,768,0.010092444717884064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,2560,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,512,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,512,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,256,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,128,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,256,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,256,0.009391999906963771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,1024,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,128,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,1024,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,64,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,32,0.0021439999755885866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,65536,0.06839111116197374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,128,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,64,0.005191111316283544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,32,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,51200,0.05487111210823059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,65536,0.05906844139099121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,256,0.002469333302643564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,65536,0.05794755617777506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,512,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,16384,0.02020355562369029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,65536,0.1330888933605618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,51200,0.04634933339224922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,51200,0.04611555404133267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,12288,0.016394666499561734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,16384,0.018211555149820115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,16384,0.018939556346999276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,51200,0.10646311442057292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,10240,0.014119999276267158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,12288,0.01570222278436025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,12288,0.014730667074521383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,768,0.002895111011134254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,16384,0.04076888826158311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,8192,0.012496000362767113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,10240,0.013975999421543546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,12288,0.03321333395110236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,10240,0.013267555170589022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,7168,0.011383111278216044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,8192,0.011993777420785693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,8192,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,6144,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,10240,0.029281778468026057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,8192,0.025661332739724055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,7168,0.010943999720944298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,5120,0.009300444689061906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,6144,0.009992889232105678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,6144,0.010092444717884064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,7168,0.023657777243190344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,4096,0.008276444342401292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,6144,0.021866665946112737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,5120,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,4096,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,4096,0.008336000144481659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,3584,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,5120,0.01958666741847992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,128,0.005662222289376789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,3072,0.006907555378145642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,3584,0.008020444048775567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,4096,0.017823111679818895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,3584,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,2560,0.005893333504597346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,3072,0.00758755538198683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,3072,0.0069955554273393415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,3584,0.015444444285498725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,2048,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,7168,0.01090666651725769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,2560,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,2560,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,3072,0.01483022173245748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,1536,0.00434311106801033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,2048,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,2048,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,1024,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,1536,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,1536,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,768,0.0028497777465316984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,2048,0.012797333300113678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,1024,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,1024,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,1536,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,512,0.002516444358560774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,768,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,1024,0.01072799993885888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,768,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,256,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,512,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,512,0.005281777845488654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,128,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,768,0.010687111152542962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,256,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,512,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,64,0.001820444439848264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,128,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,32,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,128,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,65536,0.05791911151674059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,64,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,128,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,2560,0.013810666898886362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,32,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,51200,0.04607999987072415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,5120,0.008728000024954477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,65536,0.057016889254252114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,16384,0.018224888377719455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,51200,0.04425955480999417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,51200,0.04555822081036038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,65536,0.13319733407762316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,16384,0.017589333984586928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,12288,0.01607644392384423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,51200,0.106312890847524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,256,0.004993777722120285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,16384,0.04046400056944953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,10240,0.013791999883121915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,12288,0.015108444624476962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,256,0.009316444396972656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,12288,0.033025778002209134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,10240,0.013215111361609565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,10240,0.013159111142158508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,10240,0.029185778564876978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,8192,0.011624000138706632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,65536,0.05669777923160129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,7168,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,8192,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,8192,0.025340444511837427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,6144,0.010293333066834344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,7168,0.010680889089902243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,16384,0.018896889355447557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,7168,0.023345778385798138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,6144,0.009549332989586724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,5120,0.009298666483826106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,12288,0.014447111222479077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,6144,0.021653332644038733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,4096,0.007869333028793335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,5120,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,8192,0.011194667054547204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,5120,0.019575999842749704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,4096,0.008298666940795051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,3584,0.007251555720965068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,4096,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,3584,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,3584,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,3072,0.006575999988449945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,7168,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,3584,0.015465777781274585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,2560,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,3072,0.006862222320503659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,3072,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,6144,0.009708444277445475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,3072,0.01495822270711263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,2560,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,2048,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,2560,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,1536,0.0038399998512532976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,2048,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,5120,0.008622222476535374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,2560,0.013474666410022311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,2048,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,1024,0.0031511110977994073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,1536,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,2048,0.012776888906955719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,768,0.0028773333049482773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,1024,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,1536,0.011747555600272285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,4096,0.01755555636352963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,512,0.002524444419476721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,1024,0.010703999963071613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,768,0.0052879999081293745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,768,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,256,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,512,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,768,0.009949333137936061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,512,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,128,0.0021377777059872947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,256,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,512,0.00982844415638182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,64,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,256,0.009338666995366415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,128,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,128,0.004965333475006951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,32,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,65536,0.047600888543658786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,128,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,64,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,1536,0.005879999862776862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,51200,0.03643733263015747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,65536,0.054470221201578774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,65536,0.05580000082651774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,1024,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,16384,0.01630577776167128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,51200,0.04485244552294413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,51200,0.042635556724336415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,65536,0.1330933305952284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,12288,0.013805333111021252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,16384,0.0174346665541331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,16384,0.018481777773963082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,10240,0.0124177775449223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,51200,0.10625600152545506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,12288,0.014827556080288358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,12288,0.014211555322011312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,16384,0.04065422217051188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,256,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,10240,0.012764444781674279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,12288,0.032841778463787503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,7168,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,8192,0.011386666860845355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,8192,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,8192,0.025398222936524287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,32,0.004984000076850255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,7168,0.010004444254769219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,7168,0.023372444841596816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,6144,0.009150221943855286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,5120,0.008627555436558193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,6144,0.009595555563767752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,6144,0.020632889535692003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,5120,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,4096,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,8192,0.011392888923486074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,10240,0.013292444248994192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,5120,0.018566222654448617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,10240,0.029106666644414265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,4096,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,4096,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,7168,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,6144,0.009376000199053023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,3072,0.006261333409282897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,4096,0.01683200067943997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,3584,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,3584,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,2560,0.004959110998445087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,3072,0.007479999628331926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,3072,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,2560,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,3072,0.01477688882086012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,2560,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,5120,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,2560,0.014110222458839417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,1536,0.0038524443904558816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,2048,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,2048,0.006800000038411882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,2048,0.012666666673289405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,1536,0.006337777607970768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,1024,0.003198222153716617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,1536,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,1024,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,1536,0.011781333221329583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,768,0.0028764444092909494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,1024,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,3584,0.015507555670208402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,512,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,768,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,768,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,256,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,2048,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,512,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,768,0.01036888857682546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,512,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,128,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,256,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,256,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,512,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,128,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,128,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,32,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,256,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,65536,0.03663377629386054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,3584,0.006913777854707506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,64,0.004996444616052839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,51200,0.028396444188223943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,32,0.005155555489990446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,65536,0.05462755428420173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,65536,0.05196355448828804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,1024,0.005528000079923206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,16384,0.013752000199423896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,51200,0.04105244411362542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,51200,0.044171555174721606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,12288,0.012034667034943899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,65536,0.13338400257958308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,16384,0.016692444682121277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,51200,0.10558488633897568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,10240,0.011754666765530905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,12288,0.013720889058378009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,12288,0.013807111316257052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,64,0.0019226666125986313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,8192,0.010946667028797997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,12288,0.033273776372273765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,10240,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,10240,0.012015110916561551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,7168,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,8192,0.01069777790043089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,10240,0.02869066596031189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,8192,0.010178666975763109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,6144,0.009204444785912832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,7168,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,8192,0.025171554750866357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,7168,0.010088000032636855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,5120,0.008290666672918532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,6144,0.009662222531106737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,7168,0.022650667362742957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,6144,0.008914666871229807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,4096,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,16384,0.01706133286158244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,5120,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,6144,0.020253333780500624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,5120,0.008092444803979661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,3584,0.006503110958470239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,16384,0.040736890501446195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,4096,0.008275555239783393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,3072,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,5120,0.018418666389253404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,3584,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,3584,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,4096,0.016167110866970487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,3072,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,3072,0.006755555669466655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,3584,0.015104888214005364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,128,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,2048,0.004237333519591225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,2560,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,2560,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,3072,0.014775110615624322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,1536,0.003858666867017746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,2048,0.006226666685607698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,2048,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,2560,0.013442666994200813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,1024,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,1536,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,1536,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,2048,0.012534221841229333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,768,0.0028071111689011254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,1024,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,1024,0.005631111148330901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,1536,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,512,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,4096,0.007325333025720384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,768,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,768,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,1024,0.010849778023031024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,256,0.0021520000365045336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,512,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,512,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,768,0.01035377797153261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,128,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,256,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,512,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,64,0.001806222316291597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,128,0.005624889085690181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,256,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,32,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,65536,0.025161777933438618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,128,0.007249777515729268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,64,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,32,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,51200,0.023412444525294836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,65536,0.04929600159327189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,65536,0.054422222905688815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,51200,0.03972444600529141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,65536,0.13217777676052517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,16384,0.011385777758227454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,2560,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,51200,0.1056800021065606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,12288,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,256,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,16384,0.016175111134847004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,16384,0.04058488872316148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,12288,0.013235555754767524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,128,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,10240,0.00962311112218433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,12288,0.013106666505336761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,8192,0.00869422240389718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,10240,0.012039110892348819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,12288,0.032223112053341336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,7168,0.00831911133395301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,8192,0.010665777656767102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,8192,0.009890666438473595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,10240,0.028748446040683325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,6144,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,7168,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,8192,0.02537244392765893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,51200,0.043825778696272105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,5120,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,6144,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,6144,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,7168,0.022260444031821355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,4096,0.006360000206364526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,5120,0.008707555631796518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,5120,0.00795999997191959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,6144,0.01994311147265964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,3584,0.006133333262470033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,4096,0.00795911086930169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,4096,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,5120,0.018238221605618794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,3072,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,3584,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,3584,0.007017778025733099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,4096,0.016864000095261466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,2560,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,10240,0.01149866647190518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,3072,0.007345777418878343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,3072,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,3584,0.01510933372709486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,2048,0.004466666529575984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,2560,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,2560,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,3072,0.014471999473041959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,1536,0.0038924444880750445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,2048,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,2048,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,1024,0.003176888864901331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,1536,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,2048,0.013013333082199097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,16384,0.0158595558669832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,768,0.0028159999185138275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,1536,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,1024,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,768,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,512,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,1024,0.01036977767944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,768,0.01054666688044866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,512,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,7168,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,512,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,256,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,2560,0.013667555318938361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,512,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,128,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,256,0.004944889081848992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,1536,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,256,0.009126221968068017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,128,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,64,0.001968888979819086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,128,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,32,0.001964444501532449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,65536,0.02347822156217363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,768,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,128,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,64,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,32,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,51200,0.018914666440751817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,65536,0.04871999886300829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,65536,0.05412355396482679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,16384,0.010645333263609143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,51200,0.043162667089038424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,51200,0.03921866748068068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,65536,0.13182932800716823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,12288,0.00906399968597624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,16384,0.015096000499195524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,16384,0.015835555063353646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,51200,0.105131553279029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,10240,0.00830666638082928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,256,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,12288,0.013376889129479727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,16384,0.03971644573741489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,12288,0.012464000119103326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,8192,0.007626666790909237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,1024,0.006261333409282897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,10240,0.012104888757069906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,7168,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,12288,0.03194311261177063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,8192,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,8192,0.009672000176376766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,10240,0.02792533238728841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,7168,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,7168,0.009372444616423713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,8192,0.024043555061022442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,5120,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,6144,0.009415110780133141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,7168,0.02199911077817281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,6144,0.008620444271299574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,4096,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,5120,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,5120,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,3584,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,6144,0.020288000504175823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,4096,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,4096,0.00795999997191959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,3072,0.005534222142563925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,3584,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,3584,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,10240,0.011077333655622272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,3072,0.0075786668393347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,3072,0.006654222392373615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,2560,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,2560,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,6144,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,3072,0.014440000057220459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,2048,0.004208000169859992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,2560,0.006320000108745363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,1536,0.0038773334688610504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,2048,0.006913777854707506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,2560,0.013453333742088742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,2048,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,1024,0.0030951110853089225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,1536,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,2048,0.012510221865442065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,5120,0.01825066738658481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,1024,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,768,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,4096,0.016141333513789706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,1024,0.010334221853150262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,512,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,768,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,768,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,3584,0.015557333827018738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,256,0.0022008888837363985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,768,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,512,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,512,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,128,0.0019706665641731685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,256,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,256,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,512,0.009687110781669617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,64,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,128,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,768,32,0.0018657777044508194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,1536,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,65536,0.020089776979552377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,64,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,128,0.007040888898902469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,1024,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,32,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,1536,0.011733333269755045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,51200,0.01681066718366411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,65536,0.052044444613986544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,65536,0.04812444580925835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,16384,0.009083555804358589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,51200,0.03869688842031691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,51200,0.03871999846564399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,65536,0.13157244523366293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,12288,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,16384,0.015864888827006023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,12288,0.012162666353914471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,16384,0.03973155551486545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,10240,0.00674222202764617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,128,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,12288,0.013117333253224691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,12288,0.031215998861524794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,8192,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,10240,0.011759111450778114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,10240,0.011226666470368704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,8192,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,10240,0.02755555510520935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,8192,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,7168,0.006581333362393909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,7168,0.010052444206343757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,8192,0.023741333021057978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,6144,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,7168,0.009266667068004608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,16384,0.015113777584499784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,5120,0.006581333362393909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,6144,0.009376000199053023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,6144,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,6144,0.020103999310069613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,256,0.009332444104883406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,5120,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,5120,0.01756977703836229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,4096,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,3584,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,4096,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,4096,0.01647911138004727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,3072,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,3584,0.007596444752481248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,3584,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,3584,0.015522667103343539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,3072,0.007484444313579136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,3072,0.006674666785531574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,2560,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,51200,0.10524355702930027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,2048,0.004237333519591225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,7168,0.02165599995189243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,3072,0.014185777968830533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,2560,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,5120,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,1536,0.0035546666218174826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,2048,0.006650666809744305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,4096,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,2048,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,1024,0.0029893333299292457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,2560,0.013419555293189155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,1536,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,2048,0.012383999923865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,768,0.002834666727317704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,1024,0.006178666734033161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,1536,0.005797333187527127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,512,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,768,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,1536,0.012058667010731168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,768,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,256,0.0021475555582178966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,512,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,768,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,256,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,512,0.009671111073758867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,2560,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,256,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,128,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,64,0.0018026667336622875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,128,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,32,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,128,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,64,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,1024,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,32,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,1024,0.01070755554570092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,51200,0.011382222175598145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,65536,0.0466595557000902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,512,0.004964444372389051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,65536,0.04663377669122484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,16384,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,51200,0.03773777683575948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,256,0.005547555370463266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,65536,0.131860441631741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,512,128,0.0018657777044508194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,12288,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,16384,0.014759999182489185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,16384,0.015791111522250705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,10240,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,51200,0.10439022382100423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,12288,0.013079110946920184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,12288,0.012431999875439538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,16384,0.03877777854601542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,8192,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,10240,0.011899555722872416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,12288,0.031574222776624895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,10240,0.011392888923486074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,7168,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,65536,0.013044444223244986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,8192,0.010695111420419482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,10240,0.027446221974160936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,8192,0.009716444545321995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,6144,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,7168,0.009469333622190688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,8192,0.02383288906680213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,5120,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,6144,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,6144,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,7168,0.02232444451914893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,4096,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,51200,0.03698222173584832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,6144,0.019932443896929424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,5120,0.008708444734414419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,5120,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,3584,0.005199999858935674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,4096,0.007931555310885111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,4096,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,3072,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,5120,0.01777688827779558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,3584,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,3584,0.0069884442620807225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,2560,0.005208000126812193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,4096,0.01646488904953003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,3072,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,3584,0.015663999650213454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,3072,0.006745777610275481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,2048,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,2560,0.0069208890199661255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,2560,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,3072,0.014140443669425117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,1536,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,7168,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,2048,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,2048,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,2560,0.013447110851605734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,1024,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,1536,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,2048,0.013092444174819522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,1024,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,1024,0.005431111074156231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,1536,0.01143022212717268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,512,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,768,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,1024,0.010404444403118556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,256,0.0021475555582178966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,512,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,512,0.00499555551343494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,256,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,128,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,256,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,256,0.008968888885445064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,128,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,64,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,128,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,32,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,128,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,1536,0.005704888867007361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,65536,0.011752000285519494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,256,768,0.0025555555605226094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,64,0.004863111095296012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,51200,0.010314666562610203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,65536,0.04615288972854614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,32,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,768,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,65536,0.04391377833154467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,16384,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,51200,0.037271999650531344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,768,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,51200,0.035303112533357404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,12288,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,65536,0.1304533349143134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,16384,0.014597333139843412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,16384,0.015786666009161208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,10240,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,512,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,51200,0.1044515570004781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,8192,0.005453333258628845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,12288,0.01240888900227017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,10240,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,12288,0.03155466583040025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,7168,0.005320888840489917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,8192,0.009952888720565373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,10240,0.027520888381534155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,8192,0.02369511127471924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,7168,0.010056888891590966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,6144,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,7168,0.009381333159075843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,6144,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,5120,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,7168,0.022287110487620037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,6144,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,6144,0.01977066695690155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,5120,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,12288,0.013076444466908773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,5120,0.008647110727098253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,16384,0.038731555143992104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,5120,0.017536888519922893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,3584,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,10240,0.011028444601429833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,4096,0.007979555262459649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,4096,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,3072,0.004531555705600315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,8192,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,3584,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,4096,0.016159999701711867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,2560,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,3072,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,3584,0.01550044450494978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,2560,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,2560,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,2048,0.004594666676388847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,2560,0.013401777380042605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,2048,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,1536,0.0035208890007601846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,4096,0.005163555757866966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,2048,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,2048,0.012433778080675336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,1024,0.0028782222006056043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,1536,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,1536,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,1024,0.006220444622966979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,1024,0.0052933332820733385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,1536,0.011717333561844297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,768,0.0026568888376156488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,512,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,1024,0.010370666782061258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,768,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,768,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,256,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,3072,0.014826666977670459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,512,0.005660444498062134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,512,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,128,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,768,0.01034577770365609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,256,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,512,0.009334222310119206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,256,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,64,0.001790222194459703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,128,32,0.001782222133543756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,65536,0.009002666506502364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,256,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,51200,0.008359111017651027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,128,0.006983111302057902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,16384,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,32,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,12288,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,65536,0.044159111049440175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,10240,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,51200,0.03565244542227851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,3584,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,8192,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,16384,0.014836443795098199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,7168,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,6144,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,3072,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,5120,0.00453955555955569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,8192,0.009723555710580613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,4096,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,6144,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,3584,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,5120,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,3072,0.004535111288229625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,4096,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,128,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,2560,0.004587555511130227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,2048,0.00425155543618732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,3584,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,1536,0.0035137778355015647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,3072,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,64,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,2560,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,768,0.002504000026318762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,2048,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,1536,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,512,0.002193777718477779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,256,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,1024,0.005459555735190709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,128,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,12288,0.012359999948077731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,64,0.0017706666969590718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,512,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,256,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,32,0.0018168888572189545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,65536,0.009711110757456886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,128,0.004868444469239977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,7168,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,51200,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,16384,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,64,0.0047386665311124595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,32,0.004545777622196409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,12288,0.004959110998445087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,65536,0.044440001249313354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,10240,0.005327111317051782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,51200,0.03587288988961114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,8192,0.005508444375462002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,16384,0.014753777119848462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,7168,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,12288,0.012102222277058495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,64,1024,0.0028311111446883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,6144,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,10240,0.011420444481902652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,5120,0.004559111264016894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,4096,0.004229333251714706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,8192,0.009999111294746399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,3584,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,7168,0.009293333523803288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,3072,0.0041991112132867174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,6144,0.008359111017651027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,5120,0.008033778104517195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,2560,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,4096,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,768,0.005375111268626318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,1536,0.0034968890249729156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,3584,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,10240,0.011062222222487131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,1024,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,3072,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,768,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,512,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,2560,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,256,0.0021537778278191886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,1536,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,128,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,768,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,64,0.0017973333597183228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,256,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,32,0.0018026667336622875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,128,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,64,0.004841777599520153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,32,0.004975111120276981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,48,32,2048,0.003924444317817688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,51200,0.4934408929612901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,2048,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,16384,0.3644595675998264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,1024,0.005434666656785541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,512,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,51200,1.1251049041748047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,12288,0.27553510665893555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,16384,0.16529599825541177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,12288,0.12349066469404434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,16384,0.15215199523501927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,8192,0.1862471103668213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,12288,0.09970311323801677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,7168,0.16245867146386042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,8192,0.06987555821736653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,10240,0.09110222260157268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,10240,0.10462755627102321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,7168,0.07455199956893921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,6144,0.13921954896714953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,5120,0.11656622091929118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,7168,0.061583108372158475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,6144,0.05407822132110596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,4096,0.09482400284873115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,5120,0.0555902189678616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,5120,0.05806755357318454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,3584,0.0828275548087226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,4096,0.04640444450908237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,3072,0.07185866435368855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,10240,0.22789332601759169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,3584,0.04263822237650553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,3584,0.0492524438434177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,2560,0.06046399805280897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,51200,0.3958017826080322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,8192,0.08908977773454453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,2048,0.04821066723929512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,2560,0.02681244413057963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,2560,0.03643911083539327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,1536,0.038542220989863075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,3072,0.030498666895760432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,2048,0.022291556000709534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,6144,0.06567111280229357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,2048,0.0303457776705424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,1536,0.018585777944988675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,768,0.020287111401557922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,1536,0.02397866712676154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,1024,0.01479111115137736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,512,0.013481777575280933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,1024,0.0176862229903539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,768,0.013218666944238873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,4096,0.03903466794225905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,256,0.008241777618726095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,512,0.01180266671710544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,128,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,512,0.011589333415031433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,64,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,128,0.010610666539933948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,256,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,32,0.00519466648499171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,128,0.00830488900343577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,3072,0.044255110952589244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,64,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,32,0.01276266657643848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,1024,0.025590222742822435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,65536,0.5214177767435709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,768,0.01441422270403968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,256,0.010470222267839642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,65536,1.076825777689616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,51200,0.8443813323974609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,51200,0.39505865838792587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,16384,0.27393245697021484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,12288,0.20503555403815377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,16384,0.13266132937537298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,10240,0.1714222166273329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,12288,0.09958577818340725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,65536,0.39189423455132383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,12288,0.08100089099672106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,16384,0.1155066622628106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,10240,0.08433510859807332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,8192,0.07182488838831584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,51200,0.3322355482313368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,7168,0.12270932727389866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,8192,0.05611555443869697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,10240,0.0687919987572564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,7168,0.05059377683533562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,7168,0.06096088886260986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,5120,0.0883679986000061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,6144,0.0534746679994795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,4096,0.07144533263312446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,5120,0.05132177803251478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,5120,0.04527822136878967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,3584,0.06347999970118205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,4096,0.03757777810096741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,3072,0.05525955557823181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,4096,0.032285332679748535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,3584,0.03596266773011949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,3584,0.041378666957219444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,8192,0.13889866405063206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,3072,0.03551288776927524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,3072,0.025369778275489807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,2048,0.03748622205522325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,2560,0.022328888376553852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,2048,0.019097778532240126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,2560,0.03201599915822347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,2048,0.025835555460717943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,1024,0.021385777327749465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,6144,0.1050257749027676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,768,0.015481778317027621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,1536,0.01607733302646213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,1536,0.02029333346419864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,1024,0.013331555657916598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,512,0.010781333678298526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,6144,0.04411822226312426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,1024,0.014930667148696052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,256,0.006855999843941794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,768,0.011858666936556498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,128,0.0038906666967603895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,512,0.01032977799574534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,768,0.012678222523795234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,512,0.010327999790509542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,256,0.009368889033794403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,32,0.003539555602603488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,128,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,128,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,2560,0.04623022344377306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,32,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,65536,0.3620764414469401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,1536,0.02940444482697381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,65536,0.17702666918436685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,51200,0.2840924527910021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,16384,0.07877510786056519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,65536,0.15578667322794595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,51200,0.13298132684495714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,12288,0.061215996742248535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,51200,0.11497867107391357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,10240,0.05056888858477274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,16384,0.04875377813975016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,64,0.0031822222388452957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,12288,0.03712888889842563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,256,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,8192,0.04277511106597053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,10240,0.029932445949978296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,7168,0.0366542206870185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,10240,0.03212711215019226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,8192,0.02352977792421977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,6144,0.03177333209249709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,8192,0.02790844440460205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,7168,0.02148444453875224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,5120,0.027440889014138117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,7168,0.02425688836309645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,4096,0.022801778382725183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,6144,0.021780444516075984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,5120,0.022312000393867493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,5120,0.01886666648917728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,3584,0.019911110401153564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,4096,0.01585422290696038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,3072,0.01756888959142897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,3584,0.01792622274822659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,16384,0.045612444480260216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,3072,0.013377777404255338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,3584,0.017680888374646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,2560,0.014863111906581454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,12288,0.03266577919324239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,2048,0.012448888685968188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,2560,0.012333333492279053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,64,0.008793777889675563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,2560,0.01444355481200748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,1536,0.01001422190003925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,2048,0.010728889041476779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,1024,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,2048,0.012364444633324942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,6144,0.01992266707950168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,768,0.006103111224042044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,1536,0.010634666515721215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,1024,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,512,0.004295111116435793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,1024,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,768,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,4096,0.014800000521871777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,256,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,512,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,768,0.007089777953094906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,128,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,512,0.006303999986913469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,256,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,64,0.002218666589922375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,32,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,256,0.005687110953860813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,128,0.006665777828958299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,3072,0.015673778123325773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,128,0.005547555370463266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,64,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,32,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,65536,0.26602223184373647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,51200,0.2091431087917752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,16384,0.07308266560236613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,65536,0.14674845006730822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,12288,0.05712711148791843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,65536,0.13025866614447698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,51200,0.11504533555772568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,1536,0.009097778134875828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,51200,0.10206222534179688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,16384,0.037463999456829496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,10240,0.046184000041749745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,8192,0.03651377889845107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,12288,0.029321776496039495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,16384,0.04194666610823738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,7168,0.032279110617107816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,10240,0.025454221500290766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,10240,0.028021332290437486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,6144,0.02776977750990126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,8192,0.021194666624069214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,5120,0.02390844457679325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,8192,0.02427822185887231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,7168,0.01960088809331258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,7168,0.021338666478792827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,4096,0.02034222251839108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,6144,0.01774311065673828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,3584,0.017542221479945712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,5120,0.01909955508179135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,4096,0.014228444960382251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,5120,0.016525333126386006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,3072,0.015493333339691162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,4096,0.014392000105645923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,3584,0.015063999427689446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,2560,0.012850667039553324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,3584,0.015837333268589444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,3072,0.012303999728626676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,2048,0.011029333704047732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,3072,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,2560,0.011333333121405708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,1536,0.008995555341243744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,2560,0.013083555632167392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,2048,0.00964355551534229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,1024,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,768,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,1536,0.0083137775460879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,1024,0.007567110988828871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,1536,0.009740444521109263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,512,0.0042160000238153665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,1024,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,768,0.0070675553547011475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,256,0.0031679999083280563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,768,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,512,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,128,0.0028275555620590844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,256,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,512,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,64,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,32,0.0025519999778932994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,256,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,6144,0.019052444232834708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,128,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,64,0.006221333311663733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,32,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,51200,0.18180267016092935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,65536,0.12331288390689427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,16384,0.05858044491873848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,2048,0.011471110913488599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,12288,0.032320890161726214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,51200,0.09626933601167466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,16384,0.04002044598261515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,12288,0.046987556748920016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,16384,0.03566311134232415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,128,0.006568000134494569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,10240,0.03961155480808682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,12288,0.028184887435701158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,12288,0.03088800112406413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,8192,0.03385511040687561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,10240,0.025127111209763422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,65536,0.2330311139424642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,10240,0.02695111102528042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,8192,0.02163733376397027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,8192,0.023079999619060095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,6144,0.025960889127519395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,65536,0.14509333504570857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,7168,0.01891999940077464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,5120,0.021984888447655573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,6144,0.017498667041460674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,51200,0.10941866371366714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,4096,0.018453333112928603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,5120,0.017855111095640395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,6144,0.0181377778450648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,3584,0.01607022186120351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,5120,0.015858666764365304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,4096,0.014118222726715935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,3072,0.014367999302016364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,3584,0.014148443937301636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,4096,0.013935999737845527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,2560,0.01219733307758967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,3584,0.013735111388895245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,3072,0.011696000066068439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,2048,0.010642666783597736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,3072,0.012730666332774691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,2560,0.010695999695195092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,1536,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,2560,0.011642666326628791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,2048,0.009004444711738164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,1024,0.006908444480763541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,2048,0.01017599966790941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,1536,0.008310221963458592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,768,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,1536,0.008261333737108443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,1024,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,7168,0.02970577942000495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,768,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,1024,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,256,0.0032275555034478507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,768,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,128,0.0028106667515304354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,512,0.00656444455186526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,256,0.0063768890168931745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,512,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,7168,0.02014755540423923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,64,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,32,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,256,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,128,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,128,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,65536,0.17245954937405059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,64,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,32,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,51200,0.1366879940032959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,16384,0.04620000057750278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,65536,0.0926320023006863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,65536,0.06941599978340997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,12288,0.03587466809484694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,51200,0.05690311061011421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,51200,0.07096978028615315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,10240,0.02948533495267232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,12288,0.018406222263971966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,8192,0.025096888343493145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,16384,0.02783999840418498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,12288,0.02176711128817664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,10240,0.01680888897842831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,7168,0.021506667137145996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,6144,0.018929777873886954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,10240,0.019152000546455383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,512,0.0041777777175108595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,5120,0.016202666693263583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,8192,0.01680444512102339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,7168,0.013356444736321768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,4096,0.013723555538389417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,6144,0.01239822225438224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,5120,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,6144,0.013427555561065674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,3584,0.012049777640236748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,4096,0.010500444306267632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,3072,0.010794666906197866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,4096,0.010759111079904767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,3584,0.01072533345884747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,2560,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,3584,0.011051555474599203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,3072,0.008643555144468943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,2048,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,3072,0.010450666977299584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,2560,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,1536,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,2560,0.009388444324334463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,2048,0.007374222079912822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,16384,0.02320266597800785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,2048,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,1024,0.004986666556861666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,1536,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,1536,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,768,0.003903111235962974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,1024,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,512,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,8192,0.014630221658282809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,1024,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,768,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,256,0.0025128889828920364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,7168,0.014666666587193808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,768,0.00608711110221015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,512,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,128,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,64,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,256,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,512,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,32,0.0022168887986077205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,5120,0.012087110843923358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,256,0.004947555561860402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,128,0.005664888769388199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,128,0.005032000028424793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,65536,0.15830488999684653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,64,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,32,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,51200,0.12441866927676731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,16384,0.04285866684383816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,65536,0.08638666735755073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,12288,0.03362133436732822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,51200,0.06436622142791748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,16384,0.02162488963868883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,16384,0.025941333836979334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,12288,0.017789334058761597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,10240,0.028390222125583228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,8192,0.024304888314670984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,10240,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,7168,0.02046133412255181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,10240,0.017543110582563613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,8192,0.013855111267831592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,6144,0.017714665995703805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,8192,0.015503999259736804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,7168,0.013099555340078143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,5120,0.015190222197108798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,6144,0.012063999970753988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,7168,0.013746666411558786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,4096,0.013099555340078143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,6144,0.012715555727481842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,3584,0.011380444798204633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,5120,0.011052444577217102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,4096,0.009678222239017487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,5120,0.011443555355072021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,3072,0.010355555348926121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,4096,0.0105262224872907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,3584,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,2560,0.008911111288600499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,3584,0.010856889188289642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,65536,0.06651910808351305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,2048,0.007816000117195977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,2560,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,3072,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,1536,0.006341333190600078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,51200,0.05396177702479892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,2560,0.00905777762333552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,2048,0.007264888948864407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,1024,0.0045422220395671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,768,0.003891555385457145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,12288,0.020218667056825426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,2048,0.007620444728268518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,512,0.003173333282272021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,1024,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,1536,0.006691555596060223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,256,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,1024,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,768,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,128,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,512,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,768,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,64,0.0021671110557185277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,256,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,32,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,128,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,128,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,65536,0.13674933380550808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,32,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,51200,0.0958755546145969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,3072,0.00812888890504837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,65536,0.06413422028223674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,16384,0.035770667923821345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,1536,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,51200,0.05256444546911451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,51200,0.0582631097899543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,12288,0.029042667812771265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,16384,0.023744000328911677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,16384,0.02084355552991231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,10240,0.025041777226659987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,12288,0.017479999197853934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,12288,0.018175111876593698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,8192,0.020615999897321064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,512,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,10240,0.01517599986659156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,256,0.005041777673694822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,7168,0.01869066721863217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,10240,0.01607733302646213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,8192,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,6144,0.016746666696336534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,64,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,8192,0.014481777946154276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,7168,0.012696888711717395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,5120,0.014811555544535318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,7168,0.013028444515334235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,6144,0.011704888608720569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,5120,0.010440000229411654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,6144,0.012039999994966718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,3584,0.010665777656767102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,65536,0.07670755518807305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,4096,0.009162666896979014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,5120,0.010995555255148145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,3072,0.009741333623727163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,2560,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,4096,0.009949333137936061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,3584,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,2048,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,3072,0.00813155538505978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,3584,0.01034311122364468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,1536,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,3072,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,2560,0.007706666986147563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,1024,0.0038968887594011095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,2048,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,2048,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,768,0.0035137778355015647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,1024,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,1024,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,512,0.0028640000770489373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,768,0.006216889040337668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,768,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,256,0.0024844445288181305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,512,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,512,0.005488889084921942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,4096,0.011716444459226398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,256,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,256,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,64,0.0021066665649414062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,128,0.005881777654091517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,32,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,128,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,65536,0.10761777559916179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,64,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,2560,0.008328888979223039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,32,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,1536,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,51200,0.08668977684444851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,1536,0.0063546668324205614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,65536,0.06158488988876343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,65536,0.06958578030268352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,12288,0.025387555360794067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,51200,0.050104889604780406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,51200,0.05304888884226481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,16384,0.020245333512624104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,10240,0.02125866711139679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,16384,0.021528000632921856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,12288,0.01681777834892273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,8192,0.01757955551147461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,12288,0.01684533390733931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,7168,0.01588355501492818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,128,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,6144,0.014123555686738757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,10240,0.014888889259762235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,7168,0.012440000143316058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,8192,0.014096889230940076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,5120,0.014072888427310519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,6144,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,7168,0.012096888489193387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,4096,0.011513777905040316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,6144,0.011428444749779172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,5120,0.00998933364947637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,3584,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,5120,0.010263111028406356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,4096,0.008903111020723978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,3072,0.00905866672595342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,4096,0.0094035557574696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,3584,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,2560,0.007853333320882585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,3584,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,16384,0.03214844399028354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,3072,0.008618666893906063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,2048,0.006911110960774952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,2560,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,1536,0.0047582222355736625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,2560,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,1024,0.0035537779331207275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,2048,0.007167999943097432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,1536,0.00674311113026407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,768,0.0032026666320032547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,2048,0.0069919998447100324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,1536,0.006308444258239534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,512,0.0028435554769304064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,10240,0.014753777119848462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,1024,0.005966222120655908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,768,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,256,0.002521777732504739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,768,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,128,0.0022106667359670005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,512,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,256,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,64,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,128,0.005698666804366642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,32,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,128,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,65536,0.08715733554628159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,64,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,3072,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,32,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,51200,0.06953333483801948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,65536,0.0598542226685418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,65536,0.06305333640840319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,16384,0.026385777526431616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,51200,0.04843822121620178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,51200,0.04801511102252536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,12288,0.021352888809310064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,16384,0.019519110520680744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,8192,0.013204444613721637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,16384,0.019736889335844252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,10240,0.018594667315483093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,12288,0.01610488858487871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,8192,0.016168889072206285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,512,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,12288,0.015495111544926962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,7168,0.01443377799457974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,10240,0.014475555883513557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,8192,0.012727110750145383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,6144,0.011690666278203329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,10240,0.01404622197151184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,7168,0.011734222372372946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,5120,0.010092444717884064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,8192,0.013110222087966071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,4096,0.009684444301658207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,7168,0.011387555963463254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,6144,0.010225777824719747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,3584,0.00795199970404307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,5120,0.008972444468074376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,6144,0.010667555862002902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,3072,0.007231111327807109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,4096,0.008295111358165741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,5120,0.009869332942697737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,2560,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,3584,0.008020444048775567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,4096,0.008626666333940294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,2048,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,3584,0.008285333712895712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,3072,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,1536,0.004221333397759331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,2560,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,3072,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,1024,0.003499555504984326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,2048,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,2560,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,768,0.0031946665710873077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,1536,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,1024,0.0063031112982167145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,1536,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,512,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,1024,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,256,0.0026497778793176017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,768,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,256,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,512,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,128,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,512,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,64,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,256,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,32,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,256,0.005043555465009478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,128,0.005640888793600931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,128,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,65536,0.07396888732910156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,64,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,51200,0.05916533205244276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,32,0.005395555661784277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,16384,0.024080889092551336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,65536,0.06000444624159071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,65536,0.058768888314565025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,12288,0.019613333874278598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,51200,0.047213332520590894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,16384,0.019297776950730216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,10240,0.016550223032633465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,16384,0.018559111489189994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,12288,0.015857777661747403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,12288,0.015015999476114908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,8192,0.014548444085650973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,7168,0.013387555877367655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,10240,0.01350488927629259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,768,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,6144,0.012650666965378655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,8192,0.012448888685968188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,7168,0.011354666617181567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,7168,0.011046222514576383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,6144,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,5120,0.011117333339320289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,6144,0.010332444475756751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,4096,0.008905777500735389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,5120,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,5120,0.00941422250535753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,3584,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,4096,0.008295111358165741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,4096,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,3072,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,3584,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,51200,0.04690488841798571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,2560,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,3584,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,2048,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,3072,0.007762666377756331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,2048,0.005094222310516569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,10240,0.014223110344674854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,2560,0.007380444142553542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,1536,0.004121777911980947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,8192,0.012498666842778524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,2560,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,2048,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,1024,0.0032248888164758682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,768,0.0028853333658642242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,1536,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,2048,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,512,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,1024,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,768,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,1024,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,256,0.0024702221983008915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,512,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,768,0.005459555735190709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,128,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,512,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,256,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,64,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,32,0.0021475555582178966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,256,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,65536,0.06542578008439806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,64,0.00500266667869356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,32,0.005196444276306364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,51200,0.05191644363933139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,65536,0.05748533540301853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,65536,0.057200888792673744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,16384,0.020279111133681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,3072,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,51200,0.04620711008707682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,12288,0.017177777157889474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,51200,0.0450817777050866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,10240,0.014589332871966891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,16384,0.018947554959191214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,16384,0.018079110317760043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,12288,0.0159288893143336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,1536,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,8192,0.012744888663291931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,10240,0.013800000150998434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,7168,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,10240,0.013111111190583972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,8192,0.0119982221060329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,6144,0.010689777632554373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,7168,0.010517333116796283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,8192,0.011945777469211154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,5120,0.009895111123720804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,7168,0.011008888483047485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,6144,0.009665778113736046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,128,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,6144,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,5120,0.008988444175985124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,3584,0.007338666253619724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,5120,0.008952889177534316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,3072,0.006752000086837345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,4096,0.008188444707128737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,2560,0.00600888869828648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,3584,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,4096,0.007632888853549957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,2048,0.005147555636035072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,3072,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,3584,0.007270221908887227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,1536,0.0038888889054457345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,2560,0.0074035558435651995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,3072,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,1024,0.0031804444475306403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,2048,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,2560,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,768,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,1536,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,2048,0.006214222146405115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,512,0.002472888885272874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,1536,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,768,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,1024,0.005623111294375525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,12288,0.014568888478808932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,256,0.002321777865290642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,768,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,512,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,256,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,512,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,128,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,256,0.005051555732885997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,32,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,128,0.004960888789759742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,4096,0.00832711077398724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,65536,0.05587733454174466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,64,0.005203555441564984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,32,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,51200,0.04707555638419258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,65536,0.05645333396063911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,65536,0.055049777030944824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,16384,0.01865511139233907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,51200,0.045251554912990995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,16384,0.018780445059140522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,51200,0.043402665191226535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,12288,0.015789333317014907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,16384,0.017380444539917838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,10240,0.01369866645998425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,1024,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,12288,0.015465777781274585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,10240,0.013266666895813413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,12288,0.014351111319330005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,128,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,7168,0.011405333048767514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,10240,0.012832889126406776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,8192,0.011195555329322815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,8192,0.011708444191349877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,64,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,7168,0.010064889159467485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,5120,0.008947555389669206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,7168,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,6144,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,5120,0.008718222379684448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,6144,0.009300444689061906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,3584,0.007115555306275685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,5120,0.008343111309740279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,4096,0.008282666405042013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,4096,0.007568000091446771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,3072,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,3584,0.007871111234029135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,2560,0.005777777896987067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,3584,0.007297777467303806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,3072,0.007673777639865875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,2048,0.004561777744028303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,3072,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,2560,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,1536,0.003822222352027893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,2560,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,8192,0.01202311118443807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,1024,0.0032222221295038858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,2048,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,1536,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,768,0.0028799999919202593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,1536,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,6144,0.01036711119943195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,1024,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,512,0.002513777878549364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,1024,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,256,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,768,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,4096,0.008097777764002482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,768,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,128,0.0019306666735145782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,256,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,512,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,64,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,32,0.001826666709449556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,128,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,65536,0.04831022355291578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,128,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,51200,0.03836355606714884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,64,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,32,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,16384,0.016195555528004963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,65536,0.052652445104387074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,51200,0.0447635551293691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,51200,0.04170755545298258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,16384,0.01831999917825063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,2048,0.006945778098371293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,10240,0.012290666500727335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,16384,0.016866667403115165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,12288,0.014136888914638095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,8192,0.011362666885058085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,10240,0.01223288890388277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,12288,0.014085332552591959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,7168,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,10240,0.012608889076444836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,8192,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,512,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,6144,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,8192,0.01072533345884747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,7168,0.010337777435779572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,5120,0.008747555315494537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,256,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,6144,0.009416888985368941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,4096,0.007843555675612556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,5120,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,3584,0.006784889019197888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,5120,0.008240888516108194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,4096,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,4096,0.007364444434642792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,3072,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,3584,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,3584,0.007264888948864407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,2560,0.004991111242108875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,12288,0.013646221823162503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,3072,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,2048,0.004570666700601578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,3072,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,2560,0.007128888534175024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,1536,0.003843555433882607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,2560,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,2048,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,1024,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,1536,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,1536,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,768,0.0028524444335036804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,1024,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,7168,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,512,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,6144,0.008884444832801819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,768,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,256,0.0021848889688650766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,768,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,512,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,65536,0.055382221937179565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,512,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,256,0.005330666485759947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,256,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,128,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,32,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,128,0.005094222310516569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,65536,0.037110222710503474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,64,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,32,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,2048,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,51200,0.029511110650168523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,65536,0.054773333999845714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,65536,0.050664888487921826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,51200,0.044156445397271045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,1024,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,51200,0.040123555395338274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,12288,0.01199288914601008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,16384,0.01720622181892395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,10240,0.010668444136778513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,12288,0.013392888837390475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,12288,0.01371022231049008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,128,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,8192,0.009593778186374241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,10240,0.012048000262843238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,7168,0.009043555292818282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,64,0.0019040000107553268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,8192,0.010759111079904767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,6144,0.009863110880057016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,8192,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,7168,0.010049777726332346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,5120,0.0086666668454806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,7168,0.009376000199053023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,6144,0.009528888596428765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,5120,0.008714666797055138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,6144,0.008984000318580205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,3584,0.006558222075303395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,5120,0.008057778080304464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,4096,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,16384,0.01365244471364551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,3072,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,4096,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,2560,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,3584,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,3072,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,2048,0.004555555681387584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,2560,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,16384,0.016516443755891588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,1536,0.0038337777886125776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,2048,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,2048,0.00599644457300504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,1024,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,10240,0.011735111474990845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,1536,0.006406222366624408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,1536,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,768,0.002806222273243798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,1024,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,512,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,1024,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,768,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,256,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,4096,0.007240888973077138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,768,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,128,0.001889777680238088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,512,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,256,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,64,0.001790222194459703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,128,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,256,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,32,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,128,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,65536,0.028976887464523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,3584,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,64,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,51200,0.02256711158487532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,3072,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,65536,0.05434222353829277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,2560,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,16384,0.011525332927703857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,65536,0.04821866750717163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,12288,0.009319999979601966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,51200,0.04384533233112759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,51200,0.03894222113821242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,16384,0.016016888949606154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,10240,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,16384,0.015513777732849121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,12288,0.013150222599506378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,8192,0.007749333149856991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,12288,0.012436444560686747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,10240,0.012053333222866058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,7168,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,10240,0.011377777490350934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,6144,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,8192,0.009959110783206092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,7168,0.009353777600659264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,512,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,5120,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,6144,0.009324444664849175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,4096,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,5120,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,5120,0.008197333249780867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,4096,0.0083137775460879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,4096,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,3584,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,3072,0.005542222410440445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,3584,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,3072,0.0074480002125104266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,2560,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,3072,0.0068897778789202375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,8192,0.01051288925939136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,2048,0.004524444540341695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,7168,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,2560,0.00701155596309238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,1536,0.0038533334930737815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,2560,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,2048,0.006826666494210561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,2048,0.0063386667105886675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,1024,0.0031555555760860443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,6144,0.008784888519181145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,768,0.003101333354910215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,1536,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,1024,0.006222222414281633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,512,0.002492444382773505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,3584,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,1024,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,32,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,256,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,768,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,768,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,512,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,64,0.0019697778754764134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,512,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,32,0.0019475554840432273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,256,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,128,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,65536,0.023025777604844835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,64,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,51200,0.01887022289964888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,32,0.005049777941571342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,65536,0.0540559987227122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,16384,0.010050666828950247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,65536,0.04753777715894911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,51200,0.04349066813786825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,12288,0.008096000386608971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,51200,0.03831200136078729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,16384,0.015799111790127225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,10240,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,16384,0.014826666977670459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,1536,0.0064177778032090925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,12288,0.013243555194801755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,8192,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,12288,0.012311110893885294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,10240,0.011921777493423887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,7168,0.008952000074916417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,8192,0.010683555569913654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,6144,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,128,0.0018995555324686898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,7168,0.00998133338159985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,5120,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,256,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,7168,0.009321778184837764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,128,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,6144,0.009383111364311641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,4096,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,3584,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,6144,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,5120,0.008625778059164682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,4096,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,3072,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,5120,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,2560,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,3584,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,3072,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,3584,0.007166222565703922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,2048,0.004231111043029361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,2560,0.007258666886223688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,3072,0.006779555645253923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,1536,0.0036026665733920205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,2560,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,2048,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,1024,0.0032044444233179092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,1536,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,1536,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,1024,0.006228444476922353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,10240,0.011062222222487131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,1024,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,512,0.002463999928699599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,8192,0.009761778016885122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,768,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,256,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,768,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,512,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,128,0.0020684444655974707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,512,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,256,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,64,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,4096,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,128,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,32,0.0018773333479960759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,128,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,65536,0.01632000009218852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,64,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,51200,0.013776889277829064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,32,0.005129777722888523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,65536,0.04913422134187487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,16384,0.007746666669845581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,65536,0.04684622089068095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,51200,0.03859555390146043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,12288,0.009381333159075843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,51200,0.03797066542837355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,768,768,0.0028080000645584534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,10240,0.008626666333940294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,16384,0.014831999937693277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,12288,0.013110222087966071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,8192,0.0074240002367231585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,12288,0.012383999923865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,10240,0.011750222080283694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,7168,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,10240,0.011210666762457954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,8192,0.010699555277824402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,6144,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,8192,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,7168,0.009995555712117089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,256,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,5120,0.00657066661450598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,6144,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,4096,0.00657066661450598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,6144,0.008930666579140557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,5120,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,3584,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,5120,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,4096,0.008291555775536431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,2048,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,3072,0.005201777650250329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,4096,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,3584,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,2560,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,3584,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,3072,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,2048,0.00423555572827657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,1536,0.0035617777870761026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,2560,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,3072,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,1024,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,2560,0.0063075555695427795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,1536,0.00656444455186526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,2048,0.005972444597217772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,1536,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,1024,0.006069333189063602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,512,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,1024,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,768,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,256,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,768,0.005504888792832692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,7168,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,128,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,512,0.004948444250557157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,256,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,64,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,256,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,128,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,32,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,128,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,65536,0.012794666820102267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,64,0.004848000076082018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,65536,0.04636711213323805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,32,0.004600888739029567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,51200,0.018809777167108323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,16384,0.01583377851380242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,16384,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,65536,0.04399289025200737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,12288,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,2048,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,51200,0.03736089004410638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,16384,0.015830222103330824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,512,768,0.00281866660548581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,16384,0.014436443646748861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,12288,0.013122667041089801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,8192,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,12288,0.012257777982287936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,10240,0.011715555356608497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,7168,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,10240,0.011016888750924004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,8192,0.010418666733635796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,6144,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,8192,0.00983199973901113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,7168,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,5120,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,6144,0.009322666459613377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,6144,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,4096,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,5120,0.008756444685988957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,5120,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,3584,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,4096,0.008234666453467475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,4096,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,3072,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,3584,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,3584,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,10240,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,2560,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,3072,0.0074480002125104266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,2048,0.004193777839342753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,3072,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,2560,0.007195555501514011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,1536,0.003499555504984326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,512,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,2048,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,1024,0.0029884444342719186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,2048,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,1536,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,7168,0.009303111169073317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,1536,0.005670222143332164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,768,0.0026213334252436957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,1024,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,512,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,1024,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,256,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,768,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,512,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,128,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,768,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,512,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,64,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,256,32,0.0017964444640609955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,128,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,256,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,65536,0.016872000363137987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,128,0.005075555708673266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,51200,0.014821334017647637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,64,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,32,0.005103111267089844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,16384,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,65536,0.04351644383536445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,51200,0.037309332026375666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,12288,0.005519999812046687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,51200,0.03523111012246873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,2560,0.006406222366624408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,16384,0.015533333023389181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,10240,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,16384,0.014789332946141561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,8192,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,12288,0.013453333742088742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,12288,0.012206222448084088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,10240,0.011709333293967776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,7168,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,10240,0.01112711098459032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,8192,0.010732444624106089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,6144,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,5120,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,8192,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,6144,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,7168,0.009096889032257928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,4096,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,5120,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,6144,0.008681777450773451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,3584,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,5120,0.007872888611422645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,4096,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,3072,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,3584,0.007679111427730984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,51200,0.0352862212393019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,2560,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,65536,0.04618755645222134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,3072,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,2048,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,3072,0.006827555596828461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,2560,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,1536,0.003528000166018804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,2048,0.006663110935025745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,2560,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,1024,0.0028764444092909494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,2048,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,1024,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,768,0.002542222125662698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,1024,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,768,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,7168,0.00998488896422916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,512,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,768,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,256,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,512,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,512,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,128,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,256,0.005429333282841577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,256,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,4096,0.007280888656775157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,64,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,128,32,0.0017928888814316855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,3584,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,65536,0.01071999967098236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,128,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,51200,0.007248000138335758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,64,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,32,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,16384,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,1536,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,12288,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,65536,0.04338399900330437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,10240,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,1536,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,8192,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,7168,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,51200,0.03490666548411051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,6144,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,16384,0.014477334088749357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,5120,0.004543111142185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,10240,0.011066666907734342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,4096,0.004397333496146732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,7168,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,3584,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,5120,0.00794044468137953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,3072,0.004567111117972268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,4096,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,3584,0.006909333169460297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,2048,0.0042133331298828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,3072,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,1536,0.0035146665241983202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,2560,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,1024,0.0028888889484935333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,2048,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,768,0.002479111154874166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,1536,0.00565244464410676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,512,0.0022088889446523455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,12288,0.012418666647540199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,768,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,256,0.0019866666860050624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,128,0.0018782222436534034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,8192,0.009944889280531142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,64,0.00179822225537565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,6144,0.008387555678685507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,32,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,256,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,65536,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,64,2560,0.004861333303981357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,51200,0.007672888537247975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,32,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,16384,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,128,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,12288,0.004961777892377642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,65536,0.043170667356914945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,10240,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,51200,0.03490666548411051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,8192,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,16384,0.01444355481200748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,7168,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,1024,0.005326222214433882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,10240,0.01090222183201048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,6144,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,5120,0.004588444613748127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,8192,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,4096,0.004239999999602635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,7168,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,512,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,3584,0.004575110971927643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,6144,0.008623110751310984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,3072,0.004569777597983678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,5120,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,4096,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,2048,0.0042453333735466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,128,0.004980444494220945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,3584,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,1536,0.0035173334181308746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,3072,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,64,0.004951111144489712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,1024,0.002842666581273079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,768,0.002550222186578645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,2560,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,512,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,256,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,2048,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,128,0.0018666666001081467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,1536,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,1024,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,768,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,64,0.001806222316291597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,512,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,32,0.0014968889041079415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,256,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,128,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,12288,0.012106666962305704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,64,0.005021333280536863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,32,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,33,32,2560,0.00462311092350218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,16384,0.36580533451504177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,16384,0.16580711470709905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,51200,0.49281777275933164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,12288,0.27393245697021484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,51200,1.1269102096557617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,12288,0.12338044908311631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,16384,0.13565422428978816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,10240,0.2250586615668403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,10240,0.10426755746205647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,51200,0.4373226695590549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,12288,0.11168621646033393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,8192,0.18608178032769096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,10240,0.09927466842863296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,8192,0.08965955840216742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,7168,0.16185600227779812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,8192,0.07271822293599446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,12288,0.20765866173638237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,7168,0.07470044162538317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,10240,0.17922133869595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,7168,0.06419022215737237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,6144,0.13899644215901694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,5120,0.11663288540310329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,6144,0.06583199898401897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,8192,0.1432524389690823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,6144,0.05582133266660902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,4096,0.09375022517310248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,7168,0.12494044833713108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,5120,0.051516443490982056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,5120,0.05580888854132759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,3584,0.08275377750396729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,4096,0.04636266827583313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,4096,0.04030311107635498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,5120,0.09312711159388225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,6144,0.1106097764439053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,16384,0.2778764565785726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,3584,0.04610666632652283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,3072,0.07175822390450372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,3584,0.035975111855400935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,2560,0.05972533093558418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,3584,0.07705333497789171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,2560,0.036426666710111834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,3072,0.06884000036451551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,2560,0.07294311126073201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,2048,0.02937155630853441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,2048,0.02367288867632548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,4096,0.07784444755978055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,1536,0.03737866547372606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,2048,0.04976711008283827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,1536,0.023410666320059035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,3072,0.040602667464150324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,1024,0.0262755552927653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,3072,0.03154133425818549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,2048,0.048375109831492104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,1536,0.041464000940322876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,768,0.019838222199016146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,1024,0.015045333239767285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,1024,0.0173688895172543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,512,0.013813333378897773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,768,0.013468444347381592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,1024,0.032849778731664024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,768,0.014422222971916199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,256,0.00723555518521203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,51200,0.8627529144287109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,512,0.011988444460762871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,512,0.011759999725553723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,768,0.02871733241611057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,128,0.005412444472312927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,256,0.01035377797153261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,256,0.0090506664580769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,512,0.023927999867333308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,64,0.004289777742491828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,32,0.004239999999602635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,128,0.01055111073785358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,1536,0.019160888261265226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,64,0.010060444474220276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,32,0.012539555629094442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,256,0.021913778450753953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,2560,0.027850667635599773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,65536,0.519012451171875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,51200,0.3946062193976508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,51200,0.8581191168891059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,65536,1.0999049080742729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,128,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,128,0.015586665934986539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,16384,0.27848267555236816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,65536,0.4429786735110813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,16384,0.13219022750854492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,16384,0.108188443713718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,12288,0.20897066593170166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,51200,0.34560267130533856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,12288,0.0820853312810262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,10240,0.17342844274308947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,10240,0.08424444331063165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,16384,0.22471377584669325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,8192,0.14249600304497614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,10240,0.07945688565572102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,12288,0.1735173331366645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,8192,0.07182311349444918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,8192,0.0573857757780287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,12288,0.09989955690171982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,10240,0.14299733108944362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,7168,0.12368888325161403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,6144,0.105968885951572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,7168,0.06079822116427951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,7168,0.0528711113664839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,8192,0.11693867047627766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,5120,0.09015022383795844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,6144,0.04606311188803779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,6144,0.053427557150522866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,7168,0.10521688726213242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,5120,0.045455998844570585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,6144,0.09200533231099446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,5120,0.04247911108864678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,4096,0.037470221519470215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,5120,0.08017688989639282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,4096,0.033182223637898765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,3584,0.06381066640218098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,3072,0.056120885743035205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,3584,0.029432889488008287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,3072,0.033117334047953285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,3584,0.06280622217390272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,2560,0.04680711030960083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,4096,0.07326310873031616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,3072,0.025757332642873127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,3072,0.0550488895840115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,51200,0.7002124256557889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,2560,0.03141955534617106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,2048,0.037872887319988675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,2560,0.022674666510687932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,1536,0.02957333458794488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,2048,0.0196426659822464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,2048,0.02741333345572154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,2560,0.05737600061628553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,1536,0.019816888703240287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,2048,0.04161511196030511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,65536,0.8866293165418836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,1024,0.021292444732454088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,3584,0.03844622108671401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,768,0.015445333388116626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,1024,0.01593866613176134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,1536,0.0356391105386946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,512,0.010871999793582492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,1024,0.01477599971824222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,768,0.013090666797426013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,1024,0.028063999281989202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,768,0.012695111334323883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,256,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,768,0.024922667278183833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,512,0.010324444207880232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,512,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,128,0.003531555748648114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,256,0.009470221896966299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,512,0.02163022259871165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,256,0.007697777615653143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,64,0.0028435554769304064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,32,0.0031884445084465873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,128,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,256,0.018276444739765592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,128,0.014067555467287699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,64,0.009036444127559662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,32,0.010759111079904767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,1536,0.016460445192125108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,4096,0.06736622254053752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,65536,0.3590533203548855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,65536,0.1814195579952664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,51200,0.2580222288767497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,16384,0.07912444406085543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,51200,0.1330675548977322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,65536,0.16104177633921304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,12288,0.06069599919848972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,51200,0.12562755743662515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,16384,0.04861422379811605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,12288,0.036978665325376726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,128,0.009649777577983009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,12288,0.032998220788107976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,16384,0.055277334319220654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,10240,0.05085866649945577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,16384,0.043024000194337636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,8192,0.04215288824505276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,10240,0.029575109481811523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,10240,0.03209688928392198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,12288,0.04371644390953911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,7168,0.036307556761635676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,8192,0.02789777848455641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,8192,0.024725332856178284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,10240,0.03905333413018121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,6144,0.031848000155554876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,51200,0.14067199495103624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,7168,0.024081778195169237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,6144,0.021890666749742296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,65536,0.18335555659400093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,7168,0.030839999516805012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,5120,0.027318222655190363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,6144,0.027664888236257765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,5120,0.017700443665186565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,4096,0.024226667152510747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,5120,0.018817777434984844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,4096,0.016009777784347534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,5120,0.025447999437650044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,4096,0.01608888970481025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,3584,0.019879110985332064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,4096,0.021464889248212177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,3584,0.014815110299322339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,3072,0.01756000022093455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,3584,0.01716977854569753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,8192,0.03416000141037835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,3584,0.020945777495702107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,2560,0.014831999937693277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,3072,0.015509333875444202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,6144,0.020215999748971727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,2048,0.01238044434123569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,3072,0.018364444375038147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,2560,0.01236977759334776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,2560,0.013886222408877479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,1536,0.009752000371615091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,2048,0.011114666859308878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,2560,0.01719111038578881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,2048,0.012516444755925072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,1024,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,1536,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,1536,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,2048,0.015864888827006023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,768,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,1024,0.007898666792445714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,1024,0.008627555436558193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,1536,0.0148026661740409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,512,0.0041786668201287585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,768,0.007247111035717859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,768,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,1024,0.013110222087966071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,256,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,3072,0.013541333377361298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,512,0.006868444383144379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,768,0.012316444681750404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,512,0.006500444478458828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,128,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,256,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,512,0.00998933364947637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,64,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,128,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,128,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,256,0.009676444033781687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,32,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,128,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,64,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,32,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,65536,0.257006221347385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,65536,0.1500648922390408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,51200,0.20218400160471597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,16384,0.07181600067350599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,65536,0.13043199645148382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,7168,0.022224000758594934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,51200,0.10920711358388265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,12288,0.056292441156175405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,16384,0.03800444470511542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,256,0.00563822231358952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,16384,0.04654755526118808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,12288,0.03178755442301432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,12288,0.032251556714375816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,10240,0.046068443192376025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,65536,0.14675645033518472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,51200,0.11503556039598252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,10240,0.027710222535663184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,10240,0.02801333202256097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,8192,0.03826489051183065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,7168,0.03354133168856303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,8192,0.021947556071811255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,10240,0.033599111768934466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,51200,0.11516355143653022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,6144,0.02944799926545885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,7168,0.02146666745344798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,8192,0.03013955553372701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,7168,0.020289777053727042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,5120,0.021848888860808477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,16384,0.042190220620897084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,6144,0.017818666166729398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,7168,0.026659554905361597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,6144,0.019107555349667866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,4096,0.018851555056042142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,5120,0.01629155543115404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,5120,0.016208888755904306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,3584,0.015834665960735746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,6144,0.02436622149414486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,4096,0.01442488862408532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,4096,0.014747555057207743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,3072,0.014104889498816596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,5120,0.021410667233996924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,12288,0.03774311145146688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,3584,0.013408000270525614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,3584,0.015824000040690105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,2560,0.013029333617952136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,3072,0.012448888685968188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,3072,0.014030221435758801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,3584,0.018453333112928603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,8192,0.0240328891409768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,2048,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,2560,0.011459555890825061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,3072,0.0169840008020401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,2560,0.012700444294346703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,1536,0.008371555142932469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,2048,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,2048,0.011348444554540845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,2560,0.016141333513789706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,1024,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,1536,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,1536,0.008928000099129146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,768,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,2048,0.015218666858143277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,1024,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,1536,0.013853333062595792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,1024,0.007339555356237624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,512,0.00453599997692638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,768,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,768,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,256,0.0028568889117903183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,1024,0.01257333325015174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,512,0.006675555474228329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,4096,0.019425777925385367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,512,0.005966222120655908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,768,0.010392000277837118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,128,0.0025395556456512874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,256,0.006322666588756774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,512,0.009376000199053023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,256,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,32,0.002297777682542801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,128,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,128,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,256,0.009009777671760982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,64,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,32,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,65536,0.14323822657267252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,51200,0.17997422483232287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,65536,0.1282000011867947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,51200,0.10873777998818292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,16384,0.06064977910783556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,51200,0.10232088963190715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,64,0.0021893332401911416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,16384,0.03998311029540168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,65536,0.13510666953192815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,128,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,16384,0.03580088747872247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,65536,0.2306497891743978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,12288,0.0469359987311893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,16384,0.04314755400021871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,51200,0.10778222481409709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,8192,0.03369955552948846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,12288,0.028650667932298448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,10240,0.024626667300860088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,10240,0.026718222432666357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,7168,0.027460444304678176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,12288,0.03559644354714288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,10240,0.03202044301562839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,8192,0.028648889727062646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,7168,0.019160888261265226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,6144,0.022515555222829182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,7168,0.02572800053490533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,6144,0.017849778135617573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,6144,0.017293334007263184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,5120,0.01921866668595208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,6144,0.02427911096149021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,10240,0.03904533386230469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,4096,0.017301332619455125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,12288,0.030870222383075293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,5120,0.015840000576443143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,4096,0.0140684445699056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,5120,0.021073778470357258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,4096,0.013897778259383308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,3584,0.014502222339312235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,3584,0.012967999610635968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,4096,0.01923733287387424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,3072,0.012848888834317526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,7168,0.019996444384257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,3584,0.013912000589900546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,2560,0.010664888554149203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,3584,0.017869333426157635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,3072,0.011886222494973077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,3072,0.012801777985360889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,2048,0.009871111147933537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,2560,0.011368888947698804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,2560,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,3072,0.017610667480362784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,1536,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,2048,0.009155555731720394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,2560,0.016158221496476066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,2048,0.010316444767846001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,5120,0.016498666670587327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,1024,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,1536,0.009104889300134447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,1536,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,2048,0.014779556128713818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,768,0.00553688903649648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,8192,0.023131555981106226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,1024,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,1536,0.013155555559529198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,512,0.00423555572827657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,768,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,768,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,512,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,512,0.006525333142942852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,256,0.0029120000286234748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,768,0.010664888554149203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,256,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,512,0.01037333326207267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,128,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,256,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,256,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,128,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,64,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,128,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,32,0.0022444443570242985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,128,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,65536,0.17194399568769667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,8192,0.02125155594613817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,1024,0.006996444529957241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,1024,0.011424889167149862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,51200,0.12513244152069092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,65536,0.09267199701733059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,65536,0.07156444258160062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,16384,0.04563111066818237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,51200,0.07066577672958374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,51200,0.05838221973843045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,12288,0.03484444485770331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,16384,0.027789334456125896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,16384,0.023131555981106226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,65536,0.1327093309826321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,12288,0.021869333253966436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,10240,0.027058665951093037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,64,0.005624889085690181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,32,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,12288,0.03473155697186788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,8192,0.022500443789694045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,10240,0.01917866700225406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,10240,0.016520889268981088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,7168,0.019924443629052904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,8192,0.01519377695189582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,10240,0.031086222993002996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,8192,0.01643288963370853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,6144,0.017454221844673157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,7168,0.014621333943472968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,8192,0.027239999837345544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,6144,0.013478221992651621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,6144,0.012597333225939008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,7168,0.02472266720400916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,5120,0.015103111664454142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,5120,0.0120000003112687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,6144,0.022321777211295232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,4096,0.012725333372751871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,5120,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,16384,0.04310133390956455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,3584,0.011152889165613385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,4096,0.01069866700304879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,5120,0.020678222179412842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,4096,0.01887822151184082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,3584,0.01126399967405531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,3584,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,7168,0.01369511087735494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,3584,0.018226666582955253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,2560,0.008368888662921058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,3072,0.01034400032626258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,3072,0.01650311052799225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,2560,0.009707555174827576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,2560,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,2048,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,12288,0.019814221395386588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,51200,0.10675911108652751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,2560,0.015492444237073263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,1536,0.006568000134494569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,2048,0.007511999871995714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,2048,0.008383110993438298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,1024,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,4096,0.01073688848151101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,1536,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,1536,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,768,0.0037439999481042228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,3072,0.009725333087974125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,1024,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,1024,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,1536,0.012777778009573618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,512,0.0029804443733559716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,768,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,768,0.006007110906971826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,256,0.0024666666156715816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,512,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,3072,0.008859555754396651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,512,0.009967111051082611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,256,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,128,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,256,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,128,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,256,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,128,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,64,0.0021084443562560612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,32,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,128,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,2048,0.014911111858155994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,64,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,32,0.005640888793600931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,65536,0.15154311392042372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,51200,0.11957777871025933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,1024,0.01073155552148819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,512,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,768,0.010050666828950247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,16384,0.042216890388064914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,51200,0.06407644351323445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,51200,0.053865777121649853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,16384,0.02160177793767717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,65536,0.1325520012113783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,12288,0.033393777079052396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,51200,0.10607822073830499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,16384,0.042725334564844765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,10240,0.02830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,12288,0.02004888819323646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,12288,0.017817777064111497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,10240,0.015754666593339708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,12288,0.03545066714286804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,8192,0.023952888117896184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,65536,0.06649244493908353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,10240,0.031073778867721558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,8192,0.015440889530711703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,7168,0.01792888840039571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,8192,0.026787555880016748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,7168,0.013753777576817406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,6144,0.01586844523747762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,7168,0.024176888995700415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,6144,0.012738666600651212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,5120,0.013741333451535968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,6144,0.012101333174440594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,65536,0.08503378099865383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,10240,0.01756266587310367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,4096,0.011799111134476133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,6144,0.022328888376553852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,5120,0.011394666300879585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,5120,0.01110666659143236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,3584,0.0101742222905159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,8192,0.013826666606797112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,4096,0.009663999908500248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,5120,0.020297777321603563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,3072,0.008969777988062965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,3584,0.008926221893893348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,3584,0.010810666614108615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,4096,0.018505778577592637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,16384,0.025983111725913152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,7168,0.01295466638273663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,3072,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,3072,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,3584,0.01737866633468204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,2048,0.006767999794748094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,2560,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,2560,0.009316444396972656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,3072,0.016982222596804302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,1536,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,2048,0.008277333445019191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,2048,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,1024,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,2560,0.015288889408111572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,2048,0.014135110709402295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,1536,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,1536,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,768,0.003866666720973121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,1024,0.0063271112740039825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,1536,0.01203644441233741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,1024,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,512,0.002862222285734283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,768,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,768,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,1024,0.010423111418883005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,256,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,512,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,768,0.010405333505736457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,128,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,2560,0.007772444850868649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,256,0.005736888696750005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,256,0.005140444470776452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,512,0.009373333719041612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,64,0.001855111163523462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,32,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,128,0.00573244442542394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,256,0.008715555899673039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,64,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,65536,0.12322221861945258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,128,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,32,0.005623111294375525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,51200,0.09744711054695977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,65536,0.06395910845862494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,51200,0.05818577607472738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,16384,0.03599111239115397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,4096,0.010334221853150262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,51200,0.051856001218159996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,512,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,65536,0.13187021679348415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,16384,0.020749333831999037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,12288,0.029151111841201782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,51200,0.10611822207768758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,128,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,12288,0.017086222767829895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,16384,0.042287998729281955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,10240,0.024028443627887305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,12288,0.01848088867134518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,8192,0.020749333831999037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,10240,0.015208888385030957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,12288,0.034148444732030235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,65536,0.07584800322850545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,10240,0.016184000505341422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,7168,0.01822933389080895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,8192,0.0147706667582194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,10240,0.030064887470669214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,6144,0.01611911091539595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,7168,0.012639111114872826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,7168,0.012881778180599213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,8192,0.02606755495071411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,5120,0.014117333624098035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,6144,0.011703111231327057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,7168,0.02422399984465705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,16384,0.023336889015303716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,5120,0.010715555813577441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,4096,0.01072355525361167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,6144,0.022293332550260756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,4096,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,5120,0.0199626667631997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,4096,0.00976533359951443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,4096,0.018233777748213876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,3584,0.010438222024175854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,3072,0.008340444829728868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,8192,0.013619555367363824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,3584,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,3584,0.016882666283183627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,2560,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,3072,0.00796977761718962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,3072,0.009640889035330879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,6144,0.012157333393891653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,2048,0.0063982220987478895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,2560,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,2560,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,5120,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,1536,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,2560,0.015160888433456421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,2048,0.007045333584149678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,3584,0.009528000321653154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,1536,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,1536,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,1536,0.011730666789743634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,768,0.003628444340493944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,1024,0.00630044440428416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,1024,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,1024,0.010623111493057676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,512,0.0028719999310043124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,768,0.005636444522274866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,768,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,3072,0.016081778539551627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,256,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,512,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,2048,0.006994666324721442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,512,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,128,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,256,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,256,0.005656888915432825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,2048,0.013111111190583972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,1024,0.004259555704063839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,64,0.001960888918903139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,128,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,32,0.0019742221468024785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,128,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,256,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,65536,0.1093484428193834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,64,0.005342222336265776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,128,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,32,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,51200,0.08684000041749741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,768,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,65536,0.0686968896124098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,16384,0.03145510951677958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,65536,0.0618471105893453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,51200,0.05236622360017565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,51200,0.04973155591222975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,12288,0.025370667378107708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,512,0.005535111245181825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,16384,0.021588444709777832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,16384,0.020208888583713107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,10240,0.02165866725974613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,65536,0.1323946714401245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,51200,0.10599288675520156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,12288,0.01651822196112739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,8192,0.01906399925549825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,12288,0.016930666234758165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,10240,0.014857777290874057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,10240,0.015086222026083203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,12288,0.033639109796947904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,8192,0.01349422252840466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,10240,0.02940800123744541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,7168,0.012081777883900536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,7168,0.012293332980738746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,6144,0.014780445231331719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,8192,0.025756445195939805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,6144,0.011394666300879585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,7168,0.023592889308929443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,5120,0.013016888664828407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,6144,0.011418666276666852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,6144,0.02197866638501485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,5120,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,5120,0.010614222122563256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,5120,0.01974933346112569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,4096,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,3584,0.008655110994974772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,7168,0.01647644407219357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,4096,0.008607111043400234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,8192,0.013167999684810638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,4096,0.01790488925245073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,3072,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,3584,0.008223999705579545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,3584,0.00980088859796524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,2560,0.00702133360836241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,3072,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,3072,0.009026666482289633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,3584,0.017324444320466783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,2048,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,2560,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,2048,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,2560,0.01422044469250573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,1536,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,2048,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,2048,0.012740444805887012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,16384,0.04177422324816386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,1024,0.003891555385457145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,1536,0.006648889018429651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,1536,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,1024,0.006181333214044571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,1536,0.012057777908113269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,1024,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,768,0.0036195553839206696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,768,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,1024,0.010453333457310995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,512,0.0031555555760860443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,768,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,2560,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,256,0.002444444431198968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,512,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,768,0.010093332992659675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,512,0.005394666559166378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,128,0.002208000048995018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,4096,0.00977600034740236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,256,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,512,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,256,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,64,0.0018773333479960759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,32,0.0019582222319311565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,128,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,128,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,65536,0.08716177940368652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,64,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,32,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,128,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,51200,0.06904977560043335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,65536,0.06292533212237887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,65536,0.05967289209365845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,16384,0.02609511050913069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,51200,0.0479928884241316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,51200,0.048282666338814624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,12288,0.019746666153271992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,65536,0.13184977902306452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,16384,0.019076445036464267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,3072,0.016492444607946608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,51200,0.10558399889204238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,12288,0.016162667009565566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,16384,0.0416515568892161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,12288,0.01537333263291253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,8192,0.0148026661740409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,10240,0.014474666780895658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,12288,0.033350222640567355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,256,0.009714666340086196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,8192,0.01276088919904497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,8192,0.012717333104875354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,10240,0.02950577934583028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,7168,0.011700444751315646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,8192,0.025378665990299646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,6144,0.012095111111799875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,7168,0.011437333292431302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,6144,0.01036533299419615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,7168,0.02330400049686432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,5120,0.011912888950771756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,6144,0.011037333144081963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,16384,0.019664888580640156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,10240,0.017166222135225933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,4096,0.010370666782061258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,5120,0.009424000150627559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,5120,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,3584,0.00907199995385276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,4096,0.008387555678685507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,5120,0.019310222731696237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,4096,0.008967999782827165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,4096,0.017824000782436795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,3584,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,3584,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,7168,0.01330844478474723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,3584,0.01651822196112739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,3072,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,2560,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,3072,0.008023999631404877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,3072,0.01518133282661438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,2560,0.0074035558435651995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,2560,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,2560,0.014175999495718213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,6144,0.021648888786633808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,2048,0.006487999939256244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,1536,0.004509333521127701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,2048,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,2048,0.01310222182008955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,1024,0.003176888864901331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,10240,0.014058666096793281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,1536,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,1536,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,768,0.0028817777832349143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,1024,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,3072,0.008398222426573435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,1024,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,512,0.002493333278430833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,768,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,1024,0.010421333213647207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,512,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,512,0.005660444498062134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,768,0.010325333310498131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,2048,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,512,0.009389333426952362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,256,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,128,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,256,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,256,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,64,0.001889777680238088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,128,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,32,0.0019680000841617584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,128,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,1536,0.011672000090281168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,65536,0.0733484427134196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,64,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,32,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,768,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,51200,0.059531556235419385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,65536,0.05851822429233127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,65536,0.06037777662277222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,16384,0.022778666681713525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,256,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,51200,0.0466231107711792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,51200,0.04702488912476433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,12288,0.018558222386572096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,65536,0.13142044014400905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,16384,0.01924622224436866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,10240,0.015848888291252982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,16384,0.0184879998366038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,51200,0.1054311129781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,12288,0.015942222542232938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,12288,0.015160888433456421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,8192,0.013777777552604675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,16384,0.04131288992034064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,10240,0.013627555635240344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,12288,0.03316444489690993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,7168,0.012441777520709567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,8192,0.01238933371173011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,8192,0.01238844460911221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,10240,0.02956177790959676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,6144,0.011471110913488599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,7168,0.01148088855875863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,7168,0.011100444528791638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,8192,0.025214221742418077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,5120,0.011062222222487131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,6144,0.009628444082207149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,6144,0.010678222609890832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,7168,0.0237137774626414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,4096,0.009639110830095079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,5120,0.008981333010726506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,5120,0.009353777600659264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,6144,0.021867555048730638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,3584,0.008251555263996124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,4096,0.00832266691658232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,4096,0.008645333349704742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,5120,0.01958933307064904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,3072,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,3584,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,3584,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,4096,0.018188445104493033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,2560,0.006993778049945831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,128,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,3072,0.007092444433106317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,3584,0.016473778420024447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,3072,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,2048,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,10240,0.014142221874660917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,1536,0.0038613333470291565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,2560,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,3072,0.01515111161602868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,2560,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,2048,0.0069919998447100324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,1024,0.003228444399105178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,2048,0.0064035554726918536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,1536,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,2560,0.013797333670987023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,1536,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,2048,0.012719111310111152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,1024,0.006150222072998683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,1024,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,512,0.002505777817633417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,1536,0.011506666739781698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,768,0.005991111199061076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,768,0.005383111122581694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,256,0.0022497777309682635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,1024,0.010390222072601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,512,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,512,0.0052915554907586836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,768,0.010087110930018956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,128,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,256,0.005629333357016246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,512,0.009684444301658207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,64,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,128,0.005646222167544895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,128,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,256,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,65536,0.06525422467125787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,64,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,128,0.007195555501514011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,32,0.0051893335249688895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,51200,0.049661331706576876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,768,0.0028604444944196274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,65536,0.05774933099746704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,65536,0.057393776045905225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,16384,0.019745777050654095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,51200,0.04628711276584201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,51200,0.044734223021401294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,65536,0.13155999448564318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,12288,0.016414221790101793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,16384,0.01904000010755327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,51200,0.10530400276184082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,10240,0.013923555612564087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,256,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,12288,0.014514666464593677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,12288,0.01572800013754103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,16384,0.040630221366882324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,8192,0.01239733315176434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,32,0.002026666576663653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,10240,0.013788444300492605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,12288,0.033217777808507286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,10240,0.013214222258991666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,8192,0.011388444238238864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,8192,0.012039110892348819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,6144,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,7168,0.01036711119943195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,8192,0.02517866591612498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,7168,0.023415111833148535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,6144,0.010201777848932479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,6144,0.010144000252087912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,5120,0.009331555830107795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,5120,0.009529777699046666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,16384,0.018023111753993563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,4096,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,5120,0.00927466650803884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,4096,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,4096,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,4096,0.01753155555990007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,3584,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,10240,0.029063112205929224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,3072,0.007376888559924231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,3584,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,3584,0.01573955516020457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,3072,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,3072,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,6144,0.021962665849261816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,3072,0.01478844384352366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,2560,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,2560,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,5120,0.01924711134698656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,3584,0.008368000388145447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,2048,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,2048,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,2048,0.012431110772821637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,7168,0.01108266661564509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,1536,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,1024,0.003216000066863166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,1536,0.006568888823191325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,2560,0.006393777827421824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,1536,0.01146311147345437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,1024,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,768,0.002855111120475663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,1024,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,1024,0.011032000184059143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,2048,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,768,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,768,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,768,0.010052444206343757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,512,0.005730666634109285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,256,0.002191999927163124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,2560,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,1536,0.003914666672547658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,512,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,128,0.0018808889306253856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,256,0.005706666658322017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,512,0.009316444396972656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,256,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,64,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,128,0.005623111294375525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,128,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,32,0.00179822225537565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,256,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,65536,0.05790577994452583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,64,0.00499288903342353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,32,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,51200,0.04497155547142029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,65536,0.05651288562350803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,512,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,51200,0.04303466611438327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,65536,0.13174577554066977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,16384,0.018188445104493033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,51200,0.04515377680460612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,16384,0.018765333626005385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,16384,0.017320889565679762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,12288,0.014826666977670459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,16384,0.040130668216281466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,12288,0.01427111029624939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,12288,0.015156444576051501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,10240,0.012691555751694573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,12288,0.03318488929006789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,65536,0.05494577685991923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,8192,0.011104889214038849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,10240,0.013264888690577613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,10240,0.02880000074704488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,8192,0.011727111207114326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,7168,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,8192,0.011060444845093621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,8192,0.025470222036043804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,7168,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,7168,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,6144,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,7168,0.023016888234350417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,128,0.007584888901975419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,5120,0.008731555607583787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,6144,0.009972444838947719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,6144,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,4096,0.008096000386608971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,6144,0.02130577796035343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,5120,0.00868977771864997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,5120,0.008648888932334052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,10240,0.012993777791659037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,3584,0.0074755557709270054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,4096,0.00830844458606508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,4096,0.007672000262472365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,5120,0.01923644377125634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,3072,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,3584,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,7168,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,3584,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,2560,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,4096,0.016879111528396606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,3072,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,3072,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,2048,0.004548444516128964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,3584,0.015882665912310284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,2560,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,1536,0.004184000194072723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,2560,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,3072,0.01477688882086012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,2048,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,1024,0.0031911111954185697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,2048,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,2560,0.013335111240545908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,1536,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,768,0.002814222127199173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,1536,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,2048,0.012727110750145383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,1024,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,1024,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,512,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,1536,0.01142133358452055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,768,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,768,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,1024,0.010890666809346942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,512,0.005879111174080107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,512,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,768,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,128,0.0021048889805873237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,256,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,512,0.009705777797434065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,256,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,64,0.0018639999131361644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,32,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,128,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,128,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,256,0.008957333034939235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,51200,0.10490044620301987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,64,0.004951111144489712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,32,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,51200,0.03662933243645562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,65536,0.055571556091308594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,65536,0.052074667480256825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,51200,0.04485422372817993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,51200,0.04156444470087687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,16384,0.015115555789735583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,65536,0.13158844576941595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,256,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,51200,0.10504800081253052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,12288,0.011757333245542316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,16384,0.016869333055284288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,12288,0.014592889282438489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,10240,0.010855110983053843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,12288,0.01386666629049513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,65536,0.046689778566360474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,12288,0.03314133485158285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,8192,0.011724444727102915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,10240,0.012713777522246042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,10240,0.012442666623327466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,10240,0.028773334291246202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,8192,0.010828444527255164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,7168,0.010848888920413123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,8192,0.025072889195548162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,7168,0.009659555223253038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,6144,0.009838222629494136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,16384,0.016917333006858826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,7168,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,7168,0.022788445154825848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,5120,0.009022222624884712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,6144,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,16384,0.04018844498528375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,6144,0.008912888665994009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,4096,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,5120,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,5120,0.008632889224423302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,6144,0.020908445119857788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,3584,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,128,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,4096,0.007344000041484833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,4096,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,5120,0.018251554833518136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,3072,0.005889777921968036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,3584,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,3584,0.007246221933099959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,8192,0.011309333145618439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,2560,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,4096,0.01685422162214915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,3072,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,3584,0.01585155559910668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,2560,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,3072,0.014490667316648694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,2560,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,1536,0.0038364442686239877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,2048,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,2560,0.013743110828929476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,1024,0.003195555466744635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,1536,0.006362666686375936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,1536,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,2048,0.01237422227859497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,768,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,1024,0.005969777703285217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,1536,0.011719110939237805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,768,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,512,0.002483555633160803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,1024,0.010375111467308467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,768,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,512,0.005643555687533484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,768,0.010464888479974536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,3072,0.006989333364698622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,512,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,256,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,512,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,128,0.0019102222803566193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,256,0.005000888887378904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,2048,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,64,0.0018311111877361934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,256,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,32,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,128,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,65536,0.03519377774662442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,64,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,128,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,32,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,51200,0.028146667612923518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,65536,0.05036800106366476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,65536,0.05498399999406603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,51200,0.04447288976775276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,51200,0.03993777765168084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,256,0.0021582223061058256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,16384,0.01255555533700519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,2048,0.00433955548538102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,65536,0.13119999567667642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,12288,0.010412444670995077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,16384,0.017255110873116385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,51200,0.10456266668107773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,16384,0.016546666622161865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,10240,0.015160888433456421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,12288,0.013750221994188098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,12288,0.013101333545313941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,8192,0.013251555462678274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,128,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,16384,0.04062311185730828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,12288,0.0326204432381524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,10240,0.011931555966536203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,7168,0.012260444462299347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,8192,0.01072799993885888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,1024,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,8192,0.010012444522645738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,10240,0.028809779220157202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,6144,0.011122666299343109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,7168,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,8192,0.02535377773973677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,7168,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,5120,0.0099982221921285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,6144,0.009341333475377824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,6144,0.008814222282833522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,7168,0.02328799996111128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,4096,0.00888533310757743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,5120,0.008081778056091731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,5120,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,3584,0.008904000123341879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,4096,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,5120,0.018538667096032035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,4096,0.016311999824311998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,3584,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,3584,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,3072,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,10240,0.011984888878133563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,2560,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,3072,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,3584,0.015878222054905362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,3072,0.0069120000633928515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,2048,0.004974222017659081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,2560,0.007051555646790399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,3072,0.014911999305089315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,2560,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,1536,0.0038444445365005066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,2048,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,2048,0.00685422205262714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,2560,0.01344622257683012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,1024,0.0031600000543726813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,1536,0.006575999988449945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,1536,0.005969777703285217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,2048,0.012800888882742988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,768,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,4096,0.008345777789751688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,1024,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,1536,0.011346666349305047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,512,0.00250133333934678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,768,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,768,0.005373333477311664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,1024,0.01068800025516086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,256,0.0021235555824306277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,512,0.005756444401211209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,512,0.005192000004980299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,768,0.01017066670788659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,128,0.0019137778629859288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,256,0.005366222312053044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,256,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,512,0.009317333499590555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,64,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,128,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,128,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,256,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,65536,0.024996444582939148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,64,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,128,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,6144,0.02030044462945726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,51200,0.02092711130777995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,65536,0.04821777674886915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,1024,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,65536,0.054357333315743335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,16384,0.01165511127975252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,51200,0.038719111018710665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,51200,0.043818665875328906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,12288,0.009458666874302758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,16384,0.015248000621795654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,16384,0.016208888755904306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,51200,0.1046595573425293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,10240,0.008795555267069075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,16384,0.03972266780005561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,12288,0.013223999904261695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,10240,0.012051555845472546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,8192,0.007699555820888943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,12288,0.03185955683390299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,32,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,10240,0.0284222231970893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,7168,0.0074240002367231585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,8192,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,8192,0.009782222410043081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,32,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,6144,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,7168,0.01000355515215132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,8192,0.024324445260895625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,7168,0.009336888790130615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,5120,0.006167999986145232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,65536,0.13130666149987116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,7168,0.021652445197105408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,5120,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,5120,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,6144,0.02051466703414917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,12288,0.012392000191741519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,5120,0.018605333235528734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,4096,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,4096,0.007356444166766272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,10240,0.011043555206722684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,4096,0.016189333465364244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,3584,0.007232888705200619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,3072,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,3584,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,3584,0.015485333071814643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,3072,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,3072,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,2560,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,6144,0.008928000099129146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,3072,0.014206222361988492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,2560,0.00720355576939053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,2560,0.00656444455186526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,2560,0.01313777764638265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,2048,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,1536,0.0041991112132867174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,3584,0.0058542220956749385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,2048,0.006660444455014334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,2048,0.012375999655988483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,1024,0.0034888887570963967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,1536,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,1536,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,1536,0.011400000088744693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,1024,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,6144,0.009189333352777693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,1024,0.01054222219520145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,768,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,2048,0.00510044437315729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,512,0.002540444541308615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,768,0.005640000104904175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,4096,0.006121777825885349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,256,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,512,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,768,0.009733333355850643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,512,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,128,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,256,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,512,0.009674666656388177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,256,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,64,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,128,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,128,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,32,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,65536,0.02052177819940779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,1024,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,64,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,768,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,51200,0.03117155697610643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,32,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,65536,0.05404799845483568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,65536,0.047579556703567505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,16384,0.013592888911565145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,51200,0.04316799839337667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,51200,0.03830400109291077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,12288,0.00848444468445248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,65536,0.13068711757659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,16384,0.01612711118327247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,16384,0.015517334143320719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,10240,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,51200,0.1041182213359409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,12288,0.01313066648112403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,16384,0.04022577736112807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,12288,0.012278222375445895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,8192,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,10240,0.012031111452314587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,10240,0.011223110887739392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,7168,0.0069982219073507525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,12288,0.031823111904992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,8192,0.010676444404655032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,10240,0.02741066614786784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,8192,0.009690666364298927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,6144,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,7168,0.009993777506881291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,7168,0.009329777624871995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,128,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,8192,0.024505777491463557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,6144,0.009305777649084726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,5120,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,6144,0.019908444748984445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,4096,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,5120,0.018512000640233357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,4096,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,4096,0.008254222571849823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,3584,0.0057297779454125305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,4096,0.016198222835858662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,3584,0.007164444360468123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,256,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,3072,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,3584,0.015184889237085978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,5120,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,2560,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,3072,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,7168,0.021877333521842957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,3072,0.014827556080288358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,5120,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,2048,0.0041831110914548235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,2560,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,2048,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,2048,0.006018666757477655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,1536,0.003909333298603694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,2048,0.012772444221708508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,1536,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,3584,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,1024,0.0031866667171319327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,1536,0.0063528890411059065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,1536,0.011690666278203329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,3072,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,1024,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,6144,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,1024,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,2560,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,512,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,1024,0.010341333018408881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,768,0.005283555636803309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,2560,0.01311555587583118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,512,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,256,0.0021964444054497615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,512,0.005042666776312722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,512,0.009022222624884712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,256,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,128,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,256,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,256,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,128,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,64,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,128,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,32,0.0018168888572189545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,65536,0.026775999201668635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,128,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,768,768,0.0028115556471877625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,51200,0.021958221991856892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,32,0.004752000172932942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,65536,0.052892443206575185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,65536,0.04692888922161526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,768,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,16384,0.008909333083364699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,51200,0.03973689013057285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,51200,0.037568890386157565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,12288,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,65536,0.13058755132887098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,16384,0.01478488908873664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,16384,0.015855110353893705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,51200,0.10382488701078628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,10240,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,16384,0.03941422369745042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,12288,0.013184000220563678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,12288,0.012151111331250934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,8192,0.005681777579916849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,10240,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,12288,0.03186310993300544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,7168,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,8192,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,8192,0.0107342220014996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,64,0.004962666581074397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,7168,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,6144,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,7168,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,6144,0.009378666679064432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,5120,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,7168,0.021710221966107685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,6144,0.008556444611814287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,768,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,6144,0.01997599999109904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,4096,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,5120,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,5120,0.017864000466134813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,4096,0.008151111503442129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,4096,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,3584,0.005737777799367905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,10240,0.011008888483047485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,3072,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,4096,0.016414221790101793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,3584,0.007600000335110559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,10240,0.027631999717818365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,3584,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,2560,0.004987555659479565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,3072,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,8192,0.02437866727511088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,3072,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,3584,0.015152888165579902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,2048,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,2560,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,2560,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,1536,0.003569777641031477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,3072,0.01443822185198466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,2048,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,1024,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,2560,0.013433777623706393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,1536,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,1536,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,2048,0.012416889270146688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,768,0.0028373334142896864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,1024,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,1024,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,1536,0.011350221931934357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,768,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,1024,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,768,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,512,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,512,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,768,0.010382222632567087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,256,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,256,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,512,0.010041777458455827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,256,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,256,0.009572444690598382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,128,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,2048,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,64,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,128,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,128,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,5120,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,65536,0.016565332810084026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,64,0.004868444469239977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,512,0.002479111154874166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,32,0.004966222163703707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,51200,0.013907555076811047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,65536,0.04581155710750156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,51200,0.035641776190863714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,51200,0.037363555696275495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,65536,0.12952533033159044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,128,0.001812444378932317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,16384,0.014497778481907315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,12288,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,16384,0.015784889459609985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,16384,0.03870755434036255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,512,32,0.001806222316291597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,10240,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,12288,0.012096888489193387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,12288,0.013105777402718862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,8192,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,10240,0.011759999725553723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,12288,0.0310479998588562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,10240,0.011333333121405708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,7168,0.005495111147562663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,65536,0.04633511106173197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,8192,0.00980266680320104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,10240,0.027463111612531874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,8192,0.010460444622569613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,6144,0.005207111024194294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,16384,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,7168,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,5120,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,8192,0.023375110493765935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,6144,0.008618666893906063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,51200,0.10352710882822673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,4096,0.004613333278232151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,7168,0.02160177793767717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,5120,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,5120,0.008076444268226624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,6144,0.01996977792845832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,3584,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,4096,0.00795199970404307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,5120,0.01792266633775499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,4096,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,3072,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,3584,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,3584,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,4096,0.01614488826857673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,2560,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,3072,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,3584,0.015805333852767944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,2048,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,2560,0.006495999793211619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,3072,0.014127999544143677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,2560,0.013409777647919126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,7168,0.009977777798970541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,1536,0.003551111039188173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,6144,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,2048,0.012098666694429187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,1536,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,1024,0.0028400001012616684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,1536,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,1536,0.011085333095656501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,768,0.0028311111446883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,1024,0.005547555370463266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,1024,0.010403555300500657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,768,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,512,0.0024862223201327855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,768,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,2560,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,768,0.009703111317422655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,256,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,512,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,2048,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,512,0.00924711094962226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,256,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,128,0.0018568889548381169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,256,0.008629333641793992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,128,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,64,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,1024,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,128,0.005376889059940974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,128,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,65536,0.01163200040658315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,64,0.004569777597983678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,3072,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,32,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,65536,0.045957333511776395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,51200,0.00999466660949919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,512,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,2048,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,65536,0.12869244151645237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,256,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,16384,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,51200,0.03496711121665107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,16384,0.015799999237060547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,16384,0.014767999450365702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,12288,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,256,32,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,16384,0.038703110482957624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,10240,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,12288,0.01312444441848331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,12288,0.012392889294359418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,12288,0.031230221192042034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,10240,0.011955555942323474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,10240,0.011037333144081963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,65536,0.043452445003721446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,8192,0.010447999669445885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,10240,0.0277786652247111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,8192,0.010016889207892949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,7168,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,51200,0.03720088799794515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,6144,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,8192,0.02330844435426924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,7168,0.009393778112199571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,51200,0.10241511132982041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,7168,0.021597334080272253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,6144,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,6144,0.019888000355826486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,5120,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,5120,0.008622222476535374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,4096,0.004527111020353106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,8192,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,4096,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,4096,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,3584,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,4096,0.016144000821643405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,3072,0.004572444491916233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,3584,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,7168,0.01000088867213991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,3584,0.007143110864692264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,2560,0.004872888740566042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,3072,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,3584,0.015096889601813423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,3072,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,5120,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,2560,0.006261333409282897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,2560,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,3072,0.014135110709402295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,1536,0.003522666792074839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,2048,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,2048,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,2560,0.013810666898886362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,1024,0.002833777831660377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,5120,0.018552000323931377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,1536,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,2048,0.012433778080675336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,768,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,1024,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,1024,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,1536,0.01110577748881446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,512,0.0022239999638663395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,1024,0.010779555473062726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,512,0.005026666654480828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,768,0.009650666680600908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,256,0.0021297778520319196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,2048,0.00452622233165635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,512,0.010020444790522257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,256,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,256,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,256,0.00869066682126787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,1536,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,128,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,64,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,32,0.001820444439848264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,128,0.006653333289755716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,65536,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,64,0.004872888740566042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,51200,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,32,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,16384,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,768,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,12288,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,65536,0.04310755597220527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,10240,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,512,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,51200,0.03462666604253981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,8192,0.005295111073387994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,16384,0.014541332920392355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,6144,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,6144,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,12288,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,5120,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,10240,0.010797333386209277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,128,128,0.0018053334206342697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,4096,0.004272888931963178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,8192,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,3584,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,7168,0.009333333207501305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,3072,0.004206222378545337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,5120,0.00796977761718962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,4096,0.007335999773608313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,2560,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,128,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,3584,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,1536,0.0034942221310403612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,3072,0.006656888872385025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,1024,0.0028435554769304064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,2560,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,2048,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,768,0.002550222186578645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,512,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,1536,0.005719999886221356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,256,0.002151111140847206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,1024,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,128,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,768,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,64,0.001779555653532346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,7168,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,32,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,256,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,65536,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,128,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,51200,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,64,0.00452533322903845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,16384,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,32,0.004990222139490975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,6144,0.008351999852392409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,12288,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,65536,0.0429351098007626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,10240,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,8192,0.004578666554556953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,64,2048,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,7168,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,16384,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,6144,0.004523555437723796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,12288,0.012092444631788464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,5120,0.00423911131090588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,10240,0.010834666589895884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,4096,0.0041804446114434134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,8192,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,3584,0.004572444491916233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,7168,0.0090844440791342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,3072,0.004197333421972063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,6144,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,2560,0.004547555413511065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,2048,0.004239999999602635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,4096,0.007287110719415877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,1536,0.003506666670242945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,3584,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,512,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,1024,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,3072,0.0069208890199661255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,2560,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,512,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,2048,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,256,0.001983111103375753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,1536,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,1024,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,128,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,64,0.0017991111510329777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,768,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,32,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,256,0.0046951112647851305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,51200,0.03497599893146091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,128,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,64,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,32,0.004868444469239977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,5120,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,16384,0.3639146751827664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,16384,0.1636551088756985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,51200,0.48978667789035374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,32,32,768,0.0025822222232818604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,12288,0.27181866433885365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,10240,0.22377154562208387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,51200,1.1213955349392362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,512,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,16384,0.1364133358001709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,12288,0.10422933101654053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,10240,0.10338044166564941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,7168,0.16193777985042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,10240,0.10417599810494317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,8192,0.08887555864122178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,8192,0.18291733000013563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,51200,0.42347110642327207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,6144,0.13575111495123968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,5120,0.11440622806549072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,7168,0.09322222073872884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,7168,0.07967644267612033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,6144,0.054526223076714404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,6144,0.08005333609051175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,4096,0.0938017765680949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,5120,0.06753866540061103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,3584,0.08248533142937554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,12288,0.1221804486380683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,4096,0.039020445611741804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,3072,0.07178933090633817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,4096,0.05970578061209785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,3584,0.03491733471552531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,3584,0.0540960000620948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,2560,0.0588773356543647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,2048,0.04687822196218702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,3072,0.031601776679356895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,3072,0.046880000167422824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,2560,0.027428444888856676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,1536,0.036687999963760376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,8192,0.0714613331688775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,2048,0.02345866627163357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,2560,0.039445334010654025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,2048,0.0331857767370012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,1024,0.02552977701028188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,768,0.019467555814319186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,1536,0.025459556115998164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,1536,0.01925511161486308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,1024,0.015099555253982544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,768,0.01309333327743742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,1024,0.019380443625979953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,256,0.008040889269775814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,768,0.01518311103185018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,128,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,512,0.011759111450778114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,512,0.012789333032237159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,256,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,64,0.003567111161020067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,128,0.010370666782061258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,32,0.0035822221802340615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,256,0.009171555439631144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,128,0.008012444608741337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,64,0.010317333042621613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,32,0.009855111440022787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,65536,0.519594669342041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,512,0.013443555268976422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,51200,0.39161867565578884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,51200,0.8529706531100802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,16384,0.2773413393232557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,16384,0.13122488392723933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,12288,0.21031555864546034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,16384,0.12566488318973118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,12288,0.09825066725413005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,10240,0.1749040020836724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,65536,0.437100436952379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,12288,0.08484533098008896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,10240,0.08351822031868829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,10240,0.08448533217112224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,51200,0.367664893468221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,7168,0.1233182218339708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,8192,0.14166311422983804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,6144,0.10581066873338486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,65536,1.1052044762505424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,7168,0.06834310955471463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,7168,0.07226133346557617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,6144,0.06444978051715426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,5120,0.06053511301676432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,4096,0.07286133368810017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,6144,0.04553066690762838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,5120,0.06084444125493368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,5120,0.03865866528617011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,3584,0.06374310784869723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,4096,0.048788444863425366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,5120,0.08970222208235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,3072,0.055472890535990395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,3584,0.02906044324239095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,2560,0.04657333427005344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,3584,0.04438844323158264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,3072,0.026763555076387193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,3072,0.038408888710869685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,2048,0.0374257763226827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,1536,0.02930755416552226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,8192,0.0591813325881958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,8192,0.07088889016045465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,2560,0.033207999335394964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,1024,0.020814221766259935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,2048,0.019335110982259113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,2560,0.023008000519540574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,2048,0.02686311139000787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,768,0.015474667151769003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,512,0.010984888507260216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,1024,0.013403555585278405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,1536,0.020883555213610332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,1024,0.015834665960735746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,256,0.006894222150246303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,768,0.012325333224402534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,128,0.0036888888312710654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,512,0.01038933296998342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,512,0.010741333166758219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,256,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,64,0.0031555555760860443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,256,0.0075937774446275495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,128,0.009376889301670922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,128,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,4096,0.03186044428083632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,64,0.008980444735950893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,32,0.008781332936551835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,65536,0.33149154980977374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,51200,0.2590044339497884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,65536,0.17760888735453287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,65536,0.1462142202589247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,16384,0.08389511373307969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,51200,0.13179555204179552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,1536,0.019613333874278598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,12288,0.06331111325158013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,16384,0.04808444446987576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,51200,0.131221334139506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,768,0.0131217779384719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,10240,0.05207822057935926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,12288,0.0365973346763187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,16384,0.042428443829218544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,12288,0.033896889951494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,8192,0.04312800036536323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,7168,0.03806666533152262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,10240,0.03625955515437656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,8192,0.024453333682484094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,10240,0.031832890378104314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,6144,0.03305688831541274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,8192,0.027777777777777776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,7168,0.022487999664412603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,7168,0.030092444684770372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,4096,0.023893333143658105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,6144,0.019883554842736986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,5120,0.017937777770890128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,6144,0.02679466704527537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,3584,0.020764445265134174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,5120,0.024432000186708238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,4096,0.015549333559142219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,3072,0.01813333398765988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,4096,0.020599999361568026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,3584,0.014647111296653748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,2560,0.015152888165579902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,3072,0.013176889055305056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,2048,0.012701333396964602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,2560,0.01184622198343277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,3072,0.016467556357383728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,1536,0.010021333065297868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,2560,0.015104000767072042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,2048,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,1024,0.007463110817803278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,2048,0.01311911145846049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,1536,0.009688888986905416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,768,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,1536,0.010810666614108615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,1024,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,512,0.004353777815898259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,768,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,1024,0.009071110851234859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,256,0.002834666727317704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,5120,0.028456889920764502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,768,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,128,0.002453333387772242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,512,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,256,0.00665155549844106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,64,0.0022053333620230355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,256,0.0057608890864584185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,128,0.00666933341158761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,32,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,128,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,3584,0.018518222702874076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,32,0.006031999985376994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,65536,0.2548817793528239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,65536,0.14894844426049128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,51200,0.2000159952375624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,16384,0.07024088833067152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,32,0.0032275555034478507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,65536,0.13464177979363337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,51200,0.11296978261735703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,12288,0.054795556598239474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,16384,0.042972445487976074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,51200,0.10681688785552979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,10240,0.04556355542606778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,16384,0.04167911079194811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,12288,0.02941155433654785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,8192,0.038134220573637224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,12288,0.032045334577560425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,7168,0.033788445923063494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,512,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,8192,0.021566222111384075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,10240,0.027832001447677612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,6144,0.029294222593307495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,8192,0.024181332853105333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,5120,0.025786666406525507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,7168,0.019732443822754752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,7168,0.02609511050913069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,6144,0.018182223041852314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,64,0.006070222291681502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,5120,0.0162453336848153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,6144,0.02392622166209751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,3584,0.01667555504375034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,3072,0.014814222852389017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,5120,0.020774222082561918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,4096,0.014327110515700446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,2560,0.01238933371173011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,3584,0.013166222307417126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,4096,0.018604444132910836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,2048,0.010871999793582492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,3072,0.012352888782819113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,3584,0.01750666730933719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,1536,0.008637333081828224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,2560,0.01108977778090371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,3072,0.015096000499195524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,2048,0.009943111075295342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,2560,0.013751111096805997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,1024,0.006986666884687211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,1536,0.008691555923885768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,2048,0.012336889074908363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,768,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,1536,0.010185778141021729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,512,0.003879111260175705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,1024,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,768,0.006993778049945831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,256,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,768,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,512,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,128,0.0024666666156715816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,512,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,256,0.006476444502671559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,64,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,4096,0.01956533392270406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,256,0.0053680001033677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,32,0.0023617777559492323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,128,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,128,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,64,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,65536,0.21826399697197807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,32,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,65536,0.14012444019317627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,51200,0.17389067014058432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,16384,0.06010311179690891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,65536,0.12893599934048122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,12288,0.04933689037958781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,51200,0.10416799783706665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,1024,0.007373332977294922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,16384,0.03558399942186143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,10240,0.03992799917856852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,16384,0.039692444933785334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,10240,0.029674665795432195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,12288,0.02754222187731001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,12288,0.03066666589842902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,10240,0.02756799923049079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,7168,0.029422223567962646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,10240,0.026693332526418898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,8192,0.02083288960986667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,6144,0.025775111383861963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,7168,0.019135110908084445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,5120,0.021969777014520433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,7168,0.022494221727053326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,6144,0.017180444465743173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,4096,0.01737422247727712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,6144,0.020063999626371596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,5120,0.015551111764378019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,3584,0.015118221441904703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,5120,0.017661333084106445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,4096,0.013776000175211163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,3072,0.013478221992651621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,3584,0.01293244461218516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,3584,0.01445777714252472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,51200,0.10804977681901719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,3072,0.011950222154458364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,3072,0.013153778182135688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,2048,0.009670221971140968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,2560,0.010320888625250923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,8192,0.03467288944456313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,1536,0.008219555848174626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,2048,0.008991111483838823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,1024,0.006551110910044775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,2560,0.01183111137813992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,2048,0.01070844464831882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,1536,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,768,0.004973333328962326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,1536,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,1024,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,512,0.0038488888078265717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,1024,0.007300444775157505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,768,0.00675022229552269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,256,0.0028400001012616684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,512,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,128,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,256,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,256,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,4096,0.01566844516330295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,128,0.006216889040337668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,32,0.0021520000365045336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,128,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,2560,0.011511999699804517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,64,0.005817777580685086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,32,0.005628444254398346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,65536,0.1740880012512207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,51200,0.1377991173002455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,65536,0.09064000182681614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,65536,0.06955822308858235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,16384,0.050075554185443454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,51200,0.0693111088540819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,51200,0.059056891335381397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,12288,0.0359226663907369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,16384,0.02293066680431366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,768,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,16384,0.02752711044417487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,10240,0.026817777090602454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,512,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,12288,0.018545778261290658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,8192,0.02273688879277971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,12288,0.021709332863489788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,64,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,7168,0.01984177695380317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,10240,0.018977777825461496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,8192,0.0147706667582194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,8192,0.016362667083740234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,7168,0.013730666703648038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,5120,0.014952888091405233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,7168,0.017331555485725403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,6144,0.012568888564904531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,4096,0.013424889081054263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,6144,0.01633155511485206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,5120,0.011754666765530905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,3584,0.011058666639857821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,4096,0.010743111371994019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,4096,0.012666666673289405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,8192,0.022970666488011677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,3072,0.00998488896422916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,3584,0.011721777419249216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,2560,0.008652444514963362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,3072,0.008974221845467886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,3072,0.01072355525361167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,2560,0.008068444828192392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,2048,0.007615110940403408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,10240,0.01793599956565433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,1536,0.006227555374304454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,2560,0.009786666267448002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,6144,0.01741155485312144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,1024,0.004707555390066571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,1536,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,2048,0.008978666530715095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,768,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,1536,0.0075431110130416016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,1024,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,512,0.0028444443725877335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,1024,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,768,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,256,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,5120,0.014194667339324951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,768,0.006220444622966979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,128,0.0021333334346612296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,512,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,512,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,256,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,3584,0.009667555491129557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,128,0.005636444522274866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,256,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,128,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,65536,0.15088621775309244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,64,0.005450666778617435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,32,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,51200,0.11891199482811822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,65536,0.08432888984680176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,2048,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,16384,0.042153779003355235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,51200,0.06305866771274142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,12288,0.03281866510709127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,51200,0.0536462234126197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,16384,0.021423111359278362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,10240,0.027786665492587622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,16384,0.02567911148071289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,12288,0.017872000734011333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,8192,0.023707555400000677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,12288,0.020111110475328233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,10240,0.015835555063353646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,7168,0.018194667167133756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,10240,0.017418666018380057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,8192,0.014102222190962898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,6144,0.01576622161600325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,32,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,8192,0.015225778023401896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,7168,0.013070222404268054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,5120,0.01347644461525811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,7168,0.016750221451123554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,6144,0.01202133297920227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,5120,0.011075555450386472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,3584,0.010279999838935005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,6144,0.015104888214005364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,5120,0.013786666923099093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,4096,0.009884444375832876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,3072,0.009302222066455418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,4096,0.012039110892348819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,3584,0.00924355536699295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,2560,0.00816177742348777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,65536,0.06576533450020684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,3072,0.00790666706032223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,3584,0.011428444749779172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,2048,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,1536,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,3072,0.010673777924643623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,2048,0.007224000162548489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,2560,0.009687999884287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,1024,0.003991111285156674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,2048,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,1536,0.006647111227114995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,768,0.003175111073586676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,1536,0.007289778027269576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,1024,0.006452444526884291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,512,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,768,0.005980444451173146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,1024,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,256,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,768,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,512,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,128,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,512,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,256,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,64,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,256,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,32,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,128,0.005702222386995952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,128,0.004876444323195351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,65536,0.12540800041622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,64,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,32,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,51200,0.09881244765387641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,2560,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,65536,0.07580266396204631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,16384,0.03613600134849548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,65536,0.06377244657940336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,12288,0.028632889191309612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,51200,0.056812445322672524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,16384,0.02091466718249851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,51200,0.051642666260401406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,10240,0.024367999699380662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,12288,0.017362667454613578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,64,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,10240,0.015099555253982544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,8192,0.020587555236286588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,12288,0.018418666389253404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,7168,0.01849688920709822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,10240,0.016144000821643405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,4096,0.011668444507651858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,6144,0.01610844499535031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,8192,0.01371911085314221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,5120,0.014091556270917257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,7168,0.012847111456924014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,8192,0.014711999230914645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,6144,0.011707555916574268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,4096,0.011744000017642975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,6144,0.01426666643884447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,5120,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,3584,0.009392889009581672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,5120,0.012747555143303342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,4096,0.009005332986513773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,3072,0.008363555702898238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,4096,0.011659555964999728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,3584,0.008307555483447181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,2560,0.007558222446176741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,3072,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,3584,0.010826666322019366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,2048,0.006562666760550604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,2560,0.007431999676757389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,1536,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,2560,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,2048,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,16384,0.0236462222205268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,1024,0.0038328886859946777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,2048,0.007582222421964009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,768,0.0031475555151700974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,1536,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,1024,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,1536,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,512,0.0026977778308921387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,1024,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,768,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,256,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,512,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,768,0.005884444548024072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,128,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,64,0.0021377777059872947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,7168,0.015506666567590503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,32,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,128,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,128,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,65536,0.10659999979866876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,64,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,32,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,51200,0.08533510896894668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,65536,0.06153066953023275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,3072,0.010067555639478896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,65536,0.06798133585188124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,16384,0.031705776850382485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,12288,0.02593066626124912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,51200,0.05154577891031901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,16384,0.019991111424234178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,16384,0.02128000060717265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,10240,0.02165777815712823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,12288,0.016767111089494493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,8192,0.019564444820086162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,12288,0.01669066647688548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,10240,0.014984889162911309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,8192,0.013109332985348172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,7168,0.016594666573736403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,10240,0.014924445086055331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,512,0.005320888840489917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,8192,0.01372977760103014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,6144,0.014856888188256158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,256,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,7168,0.012190221912331052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,5120,0.013088888592190213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,6144,0.011110222174061669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,7168,0.01475288967291514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,4096,0.010013333625263637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,6144,0.013497778111033969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,5120,0.00983911090426975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,3584,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,5120,0.012256888879670037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,4096,0.0083137775460879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,3072,0.007816889219813878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,3584,0.008104888929261101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,4096,0.011324444578753578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,2560,0.007003555695215861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,3584,0.010606221854686737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,51200,0.04961422085762024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,3072,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,2048,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,1536,0.005714666512277391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,2560,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,3072,0.009584888815879822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,2560,0.008143111235565608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,1024,0.0038417776425679526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,2048,0.00699733363257514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,768,0.0038568890757030914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,1536,0.006754666566848755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,2048,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,1024,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,1536,0.00686488880051507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,768,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,1024,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,256,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,512,0.005893333504597346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,256,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,128,0.002143111079931259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,256,0.005740444279379315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,64,0.0019040000107553268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,128,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,256,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,32,0.002143111079931259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,128,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,65536,0.08745688862270778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,64,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,32,0.005294222384691238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,51200,0.06948088937335543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,65536,0.061299555831485324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,65536,0.059161775641971164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,16384,0.02570488883389367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,12288,0.019556444552209642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,51200,0.04733333322736952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,51200,0.047833777136272855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,16384,0.019389332996474374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,10240,0.01723911033736335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,16384,0.01943555474281311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,512,0.00282222218811512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,12288,0.01590844492117564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,8192,0.015102222561836243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,12288,0.015396444333924187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,7168,0.013099555340078143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,10240,0.014472888575659858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,10240,0.014003554979960123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,768,0.005649777750174205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,8192,0.012694222231705984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,5120,0.010018666585286459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,7168,0.011880888707107969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,512,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,8192,0.012895111408498553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,4096,0.008671110702885522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,6144,0.010099555883142684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,7168,0.014458666245142618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,3584,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,5120,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,6144,0.01238222254647149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,3072,0.007297777467303806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,4096,0.008274666965007782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,2560,0.006335111127959357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,3584,0.007684444387753804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,4096,0.01073688848151101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,3584,0.008992888861232335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,2048,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,3072,0.00832266691658232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,2560,0.007015110717879401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,1536,0.003999111139112049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,2048,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,2560,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,2048,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,1536,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,768,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,1536,0.006370666540331311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,1024,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,512,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,1024,0.005623111294375525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,768,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,256,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,768,0.005680000202523337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,6144,0.011521778172916837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,512,0.00564533347884814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,128,0.002161777681774563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,512,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,64,0.001971555459830496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,256,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,128,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,32,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,128,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,5120,0.01130488928821352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,65536,0.07575555642445882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,64,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,3072,0.007703111403518253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,51200,0.06017688910166422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,32,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,16384,0.02329066726896498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,65536,0.058168888092041016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,65536,0.058168888092041016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,12288,0.01863822175396813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,1024,0.0032595555401510666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,51200,0.04695466823048062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,10240,0.015807999504937064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,51200,0.04535733328925239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,16384,0.019253333409627277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,16384,0.01848088867134518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,12288,0.01583377851380242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,7168,0.01295466638273663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,10240,0.014304889572991265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,12288,0.015051555302408008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,8192,0.012432888978057437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,10240,0.013451555536852943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,8192,0.012352888782819113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,7168,0.011056889262464313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,5120,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,7168,0.013367111484209696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,6144,0.009687110781669617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,4096,0.008727110922336578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,6144,0.011697777443461947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,3584,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,5120,0.011352000137170156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,4096,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,3072,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,4096,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,3584,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,2560,0.006049777898523543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,3584,0.008407110969225565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,3072,0.007719111111429002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,2048,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,3072,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,2560,0.007053333024183909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,1536,0.0038488888078265717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,2560,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,2048,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,1024,0.0031911111954185697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,2048,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,8192,0.013693333499961428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,768,0.0028640000770489373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,1536,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,1536,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,1024,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,6144,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,768,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,1024,0.005673777725961473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,256,0.0022373333987262514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,768,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,512,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,128,0.0021475555582178966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,512,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,5120,0.008871999879678091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,64,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,32,0.001955555544959174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,256,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,65536,0.06603199905819364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,128,0.004960000101062986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,64,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,32,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,51200,0.05105333195792305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,65536,0.057327999009026416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,65536,0.0560435586505466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,16384,0.019653333557976615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,12288,0.01665599975321028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,51200,0.046128001477983266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,51200,0.043509334325790405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,16384,0.018915555543369718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,16384,0.017789334058761597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,12288,0.015594666202863058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,512,0.002537777854336633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,12288,0.0144195556640625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,10240,0.014130666851997375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,10240,0.01311911145846049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,7168,0.011400000088744693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,8192,0.011539555258221097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,6144,0.01036711119943195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,8192,0.011786667009194693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,7168,0.011049778097205691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,128,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,5120,0.009471110999584198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,7168,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,4096,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,6144,0.011027555498811932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,5120,0.008976000050703684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,3584,0.008266666697131263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,5120,0.009617778162161509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,4096,0.008307555483447181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,4096,0.007830222447713217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,3072,0.007161777880456712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,3584,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,3584,0.007515555454625024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,10240,0.014259555273585849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,3072,0.007080000307824876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,8192,0.012318222059143914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,2048,0.005486222190989389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,2560,0.007256000406212277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,2560,0.006688888702127669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,1536,0.0048142220411035745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,2048,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,256,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,1024,0.0038444445365005066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,2048,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,1536,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,1536,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,6144,0.009360888765917884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,1024,0.005980444451173146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,512,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,1024,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,768,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,256,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,768,0.005447111195988125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,512,0.0058515556156635284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,128,0.0019422223170598347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,512,0.005203555441564984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,256,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,256,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,128,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,2560,0.006393777827421824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,32,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,3072,0.007666666474607255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,65536,0.05901155869166056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,51200,0.045455111397637256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,64,0.004941333499219683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,32,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,65536,0.056399113602108426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,16384,0.01993777851263682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,65536,0.053189333942201406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,51200,0.04525244567129347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,12288,0.01608000033431583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,16384,0.018769777483410306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,51200,0.04208533300293816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,10240,0.014319111903508505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,768,0.003532444437344869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,16384,0.017320000463061862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,8192,0.01311644415060679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,12288,0.015152888165579902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,7168,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,12288,0.014142221874660917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,8192,0.010866666833559671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,6144,0.01051288925939136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,8192,0.011623111036088733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,7168,0.011029333704047732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,6144,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,5120,0.008957333034939235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,64,0.0018666666001081467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,4096,0.008023111356629265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,5120,0.00906399968597624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,3584,0.007133333219422235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,4096,0.008031999899281396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,5120,0.008616000413894653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,3072,0.007537778053018782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,4096,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,3584,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,2560,0.0069022224181228215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,3584,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,3072,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,2048,0.004990222139490975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,3072,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,2560,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,1536,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,2560,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,2048,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,10240,0.013079110946920184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,1024,0.003539555602603488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,2048,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,1536,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,10240,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,7168,0.010262221925788457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,512,0.0026613333159022858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,1536,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,1024,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,768,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,256,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,512,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,6144,0.009684444301658207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,128,0.0021226666867733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,256,0.005639111002286275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,64,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,256,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,128,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,32,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,128,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,65536,0.04769066638416714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,64,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,32,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,65536,0.055424888928731285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,51200,0.037011557155185275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,65536,0.05096888873312208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,16384,0.014742222097184924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,51200,0.04485955503251818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,768,0.0031555555760860443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,51200,0.040631999572118126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,1024,0.006231110956933763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,12288,0.013563555147912769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,10240,0.012312889099121094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,16384,0.01833511061138577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,12288,0.014136888914638095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,768,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,10240,0.012227555943859948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,512,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,10240,0.012544000314341651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,7168,0.010053333308961656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,8192,0.01073333372672399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,6144,0.009317333499590555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,8192,0.011018666956159802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,7168,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,5120,0.008600888980759515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,7168,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,6144,0.009710222482681274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,4096,0.007511110769377814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,6144,0.008958222137557136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,5120,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,3584,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,5120,0.008294222255547842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,4096,0.008294222255547842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,3072,0.005674666828579373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,4096,0.007667555577225155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,3584,0.007722666694058313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,2560,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,3584,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,3072,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,16384,0.016795555750528973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,2048,0.004318222403526306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,3072,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,1536,0.0038399998512532976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,8192,0.011377777490350934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,12288,0.013763555222087435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,2048,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,1024,0.0032071111102898917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,2048,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,1536,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,768,0.0028320000403457214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,1536,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,1024,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,512,0.0024764444679021835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,1024,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,768,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,512,0.0057591112951437635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,768,0.005304000029961268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,128,0.0019537777536445195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,512,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,256,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,64,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,256,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,128,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,32,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,128,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,65536,0.036192887359195285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,64,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,51200,0.028468443287743464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,65536,0.05484444565243191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,32,0.005003555367390315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,16384,0.012636444634861417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,65536,0.0491226679748959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,12288,0.010126222338941362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,51200,0.04416444566514757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,2560,0.0064204442832205035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,10240,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,51200,0.03906311260329352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,16384,0.016336000627941556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,12288,0.013435555828942193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,8192,0.014477334088749357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,12288,0.01348888874053955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,10240,0.012053333222866058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,7168,0.009684444301658207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,10240,0.011767110890812345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,256,0.0021848889688650766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,8192,0.010687111152542962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,6144,0.00902844468752543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,5120,0.008453333543406593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,8192,0.0099973330895106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,7168,0.010189333723651038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,4096,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,6144,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,7168,0.009394666386975182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,5120,0.008984000318580205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,3584,0.007124444676770105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,6144,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,3072,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,5120,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,4096,0.008016000191370646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,2560,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,3584,0.007899555895063613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,4096,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,3072,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,3584,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,2048,0.004301333179076512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,2560,0.007258666886223688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,3072,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,1536,0.003670222229427762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,2560,0.006516444600290722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,2048,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,1024,0.0031635556370019913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,1536,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,2048,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,768,0.002838222309947014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,1024,0.006206222292449739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,1536,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,512,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,1024,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,768,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,256,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,512,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,768,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,128,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,512,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,256,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,64,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,32,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,128,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,256,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,65536,0.028237332900365193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,128,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,51200,0.02207999924818675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,64,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,32,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,65536,0.05437511205673218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,65536,0.04715822140375773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,51200,0.04378755556212532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,12288,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,51200,0.03781066669358148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,16384,0.015841777125994366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,10240,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,16384,0.015414221419228448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,12288,0.013150222599506378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,8192,0.007600000335110559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,16384,0.01657777859105004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,12288,0.012441777520709567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,10240,0.011911999848153857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,7168,0.00699733363257514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,10240,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,8192,0.010658666491508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,6144,0.0064693333374129395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,7168,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,8192,0.009825777676370408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,5120,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,7168,0.009376889301670922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,6144,0.009319110876984065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,4096,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,5120,0.008968888885445064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,3584,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,5120,0.008048888709810045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,4096,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,3072,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,4096,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,3584,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,2560,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,16384,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,3584,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,2048,0.004846222284767362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,3072,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,2560,0.00685066646999783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,1536,0.004279110994603899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,2560,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,2560,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,2048,0.006657777975002925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,1024,0.003859555555714501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,2048,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,768,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,1536,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,1024,0.006146666490369373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,512,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,1536,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,768,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,256,0.0021848889688650766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,1024,0.00563822231358952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,128,0.002128888956374592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,768,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,512,0.005790222022268508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,64,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,256,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,32,0.0018800000349680583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,256,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,6144,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,65536,0.021987555755509272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,51200,0.017864889568752713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,64,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,32,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,65536,0.054309331708484225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,16384,0.009692444569534725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,65536,0.04654577705595228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,51200,0.04349422123697069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,12288,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,3072,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,16384,0.015671999918089975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,51200,0.037350220812691584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,10240,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,16384,0.01477777792347802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,8192,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,12288,0.013085333009560903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,12288,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,7168,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,10240,0.011054221954610614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,8192,0.010527110762066312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,6144,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,8192,0.00981866651111179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,7168,0.010029333333174387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,5120,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,7168,0.009313777916961247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,6144,0.009326222042242685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,4096,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,6144,0.008628444539176093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,128,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,5120,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,3584,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,5120,0.008035555481910706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,4096,0.00792622235086229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,3072,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,3584,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,2560,0.004867555366622077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,3072,0.007347555624114142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,3584,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,2048,0.0041804446114434134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,3072,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,2560,0.006983111302057902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,1536,0.0036755556033717263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,2048,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,2048,0.006235555642180973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,10240,0.011722666521867117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,1024,0.0031662223239739737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,1536,0.006235555642180973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,768,0.00282577777074443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,1536,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,1024,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,512,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,1024,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,768,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,512,0.005290666802061929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,256,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,512,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,128,0.001975999938117133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,256,0.0052888890107472735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,4096,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,256,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,64,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,768,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,128,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,65536,0.016169778174824186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,2560,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,32,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,51200,0.013407111167907715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,65536,0.04552888870239258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,65536,0.04821422364976671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,16384,0.011075555450386472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,51200,0.03849511014090644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,768,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,51200,0.0369120008415646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,12288,0.006694222076071634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,16384,0.014446222119861178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,10240,0.006178666734033161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,12288,0.01311555587583118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,8192,0.005640000104904175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,10240,0.011711111499203576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,12288,0.012096000214417776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,7168,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,10240,0.011046222514576383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,8192,0.010675555302037133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,128,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,6144,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,8192,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,7168,0.009995555712117089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,5120,0.0063484443558586975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,7168,0.009328000247478485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,4096,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,6144,0.009295110901196798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,64,0.005197333378924264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,5120,0.00865955568022198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,3584,0.00564533347884814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,5120,0.00830488900343577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,4096,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,3072,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,3584,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,2560,0.004862222406599256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,3584,0.006911110960774952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,3072,0.007335999773608313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,2048,0.004192000048028098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,3072,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,2560,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,1536,0.003565333369705412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,16384,0.015601777368121676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,2560,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,2048,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,1024,0.0028488888508743713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,512,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,2048,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,1536,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,1024,0.006214222146405115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,1536,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,1024,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,768,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,768,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,256,0.0021928888228204516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,512,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,6144,0.008634666601816813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,128,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,256,0.005543111099137201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,256,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,64,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,128,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,32,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,64,0.004587555511130227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,65536,0.01685244507259793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,32,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,768,0.002809777855873108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,51200,0.014471999473041959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,65536,0.04611466659439934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,65536,0.04247555467817518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,16384,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,51200,0.03707199957635667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,51200,0.03502666619088914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,12288,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,512,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,16384,0.014416888356208801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,16384,0.015793777174419828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,10240,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,12288,0.01309866706530253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,8192,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,4096,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,12288,0.012236444486512078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,10240,0.011674666570292579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,7168,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,6144,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,128,0.00452977791428566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,8192,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,7168,0.00979555563794242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,8192,0.009673777553770278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,7168,0.009336888790130615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,6144,0.009327111144860586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,4096,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,6144,0.008276444342401292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,3584,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,5120,0.008623999853928884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,5120,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,4096,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,3072,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,3584,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,4096,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,512,512,0.002504000026318762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,2560,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,2048,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,3072,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,2560,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,3072,0.006767999794748094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,1536,0.003551111039188173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,2048,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,2560,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,1024,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,2048,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,1536,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,768,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,1024,0.006039111150635614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,1536,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,512,0.0026835555003749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,768,0.005721777677536011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,1024,0.005323555734422471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,256,0.002116444417172008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,10240,0.011009777585665384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,128,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,5120,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,256,0.005416000054942236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,512,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,256,32,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,256,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,128,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,65536,0.011386666860845355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,128,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,51200,0.010096888575288985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,64,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,65536,0.045828445090187915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,32,0.004867555366622077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,16384,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,65536,0.04250133368704054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,3584,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,51200,0.03700977894994948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,12288,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,10240,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,16384,0.015581332974963717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,51200,0.033956444925732084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,8192,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,16384,0.014278221461508008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,12288,0.01310400002532535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,7168,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,10240,0.011443555355072021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,10240,0.011065777805116443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,8192,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,6144,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,7168,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,512,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,5120,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,6144,0.009303111169073317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,6144,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,4096,0.004955555415815777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,5120,0.008625778059164682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,5120,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,3584,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,4096,0.007882666256692674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,4096,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,3072,0.004561777744028303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,3584,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,3072,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,12288,0.012019555601808759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,2560,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,3072,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,8192,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,2048,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,2560,0.006907555378145642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,7168,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,1536,0.0034942221310403612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,2048,0.006761777732107375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,2560,0.0064791109826829694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,1024,0.0028782222006056043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,2048,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,768,0.0024764444679021835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,1536,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,1024,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,1536,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,768,0.0058853332367208265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,256,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,768,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,512,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,128,0.0018088887963030075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,512,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,256,0.005376889059940974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,64,0.0017848888205157386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,256,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,32,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,128,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,3584,0.007045333584149678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,65536,0.008989333278603023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,51200,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,128,0.004543111142185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,16384,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,64,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,12288,0.005207111024194294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,32,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,65536,0.04184622234768338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,51200,0.034240000777774386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,16384,0.014087110757827759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,8192,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,768,0.0053031109273433685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,12288,0.0120319997270902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,6144,0.004859555512666702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,10240,0.01071999967098236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,5120,0.004563555535342959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,128,512,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,8192,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,1024,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,7168,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,3584,0.0046880000995265115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,3072,0.004514666481150521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,6144,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,2560,0.004507555729813046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,5120,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,2048,0.003951999876234266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,4096,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,3584,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,1536,0.0032151111712058387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,1024,0.0028284444577164124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,3072,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,768,0.002516444358560774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,2048,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,512,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,1536,0.005665777872006099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,1024,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,768,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,10240,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,128,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,64,0.0017911110901170308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,512,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,7168,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,256,0.004595555365085602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,65536,0.008292444050312042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,128,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,51200,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,64,0.004862222406599256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,16384,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,4096,0.0041955556306574075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,32,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,12288,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,10240,0.005541333307822545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,65536,0.041759110159344144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,8192,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,51200,0.034267554680506386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,16384,0.014371555712487964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,7168,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,12288,0.012266666524940066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,6144,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,10240,0.011004444625642566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,2560,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,5120,0.004555555681387584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,4096,0.0041991112132867174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,7168,0.009029332962301042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,3584,0.0045422220395671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,256,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,6144,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,3072,0.004260444392760594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,2560,0.004560889055331548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,4096,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,2048,0.004513777792453766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,64,32,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,3584,0.00675733346078131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,3072,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,1536,0.0032026666320032547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,1024,0.002838222309947014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,2560,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,2048,0.006175111151403851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,512,0.002201777779393726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,1536,0.005738666488064661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,1024,0.005281777845488654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,256,0.0021199999998013177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,128,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,768,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,64,0.0018079999006456798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,512,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,32,0.0017848888205157386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,256,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,128,0.004579555657174853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,64,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,8192,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,32,0.004629333400064045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,5120,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,16384,0.32417334450615776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,17,32,768,0.0024666666156715816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,51200,0.4896720250447591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,16384,0.1635271045896742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,51200,0.9665262434217664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,12288,0.22665601306491426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,12288,0.12197421656714545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,16384,0.13740000459882948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,10240,0.18851289484235975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,12288,0.12630577882130942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,10240,0.10328266355726455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,51200,0.4239804479810927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,8192,0.15431288878122965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,8192,0.08820622497134739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,8192,0.07170489099290636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,7168,0.1329422261979845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,7168,0.09548711114459568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,8192,0.14025155703226724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,16384,0.27502844068739146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,12288,0.20572800106472441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,7168,0.06622755527496338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,6144,0.11459733380211724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,5120,0.09626489215426975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,6144,0.07785333527459039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,6144,0.055664002895355225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,7168,0.14103378189934626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,10240,0.09488177961773342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,4096,0.07945599820878771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,5120,0.06941422488954332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,5120,0.04723644587728712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,3584,0.06849510802163018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,4096,0.05961866511238945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,4096,0.038747555679745145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,5120,0.11437511444091797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,3584,0.03472888800832961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,4096,0.08930399682786729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,10240,0.1727048953374227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,3072,0.05939022037718031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,3584,0.0532915559079912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,2560,0.050048887729644775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,3584,0.08398044771618313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,3072,0.04570133487383524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,3072,0.031173331869973078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,2048,0.040954666005240545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,2560,0.03937600056330363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,2560,0.027043556173642475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,1536,0.03162311183081733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,2048,0.03321244319279989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,2048,0.028181334336598713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,2560,0.05762844615512424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,1024,0.02228533393806881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,2048,0.05043111244837443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,1536,0.02584444483121236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,1536,0.040989332728915744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,1024,0.018823999497625563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,1024,0.015118221441904703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,1024,0.03273422188229031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,768,0.01647111111217075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,512,0.012406222522258759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,768,0.01313688854376475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,768,0.02811644474665324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,3072,0.06681955522961087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,512,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,256,0.007168889045715332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,51200,0.8349359830220541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,512,0.02437155610985226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,1536,0.01960088809331258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,6144,0.12184888786739773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,128,0.005196444276306364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,256,0.009594666461149851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,256,0.010592000352011787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,64,0.0042444442709287005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,768,0.017839110559887357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,128,0.010602666272057427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,32,0.004184000194072723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,64,0.010363555616802638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,256,0.02090222140153249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,32,0.0103493332862854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,512,0.012528888881206512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,65536,0.5161315600077311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,65536,1.087110201517741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,128,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,128,0.015394666128688388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,51200,0.39087287584940594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,51200,0.8486008644104004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,16384,0.1311066680484348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,65536,0.43220088216993546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,12288,0.206441773308648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,12288,0.098380446434021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,51200,0.32494844330681694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,10240,0.17269333203633627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,16384,0.2769413259294298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,10240,0.08321422338485718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,10240,0.07629777987798055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,16384,0.22052444352043998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,12288,0.16821777820587158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,8192,0.14064889483981663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,16384,0.1089751124382019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,8192,0.07056533628039889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,10240,0.14292711681789821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,8192,0.05760800176196628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,12288,0.09255822499593098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,7168,0.07322666380140516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,7168,0.0513955553372701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,8192,0.11612533198462592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,6144,0.06435288985570271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,6144,0.046163555648591786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,5120,0.08922488821877374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,6144,0.09988800022337173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,5120,0.05997244517008463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,7168,0.1223271158006456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,4096,0.07311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,5120,0.03834933373663161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,5120,0.09432355562845866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,6144,0.10577422380447388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,4096,0.03234666585922241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,3584,0.06351733207702637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,4096,0.04859911070929634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,3584,0.04215111003981697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,3072,0.0551528897550371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,3584,0.028712888558705647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,7168,0.11764532989925808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,2560,0.046078221665488354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,3584,0.062414222293429904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,3072,0.02556444373395708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,3072,0.038091556893454656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,51200,0.6924346817864312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,2048,0.03692266676161025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,2560,0.02239466706911723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,65536,0.894095950656467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,3072,0.055249777105119496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,1536,0.029701332251230877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,2048,0.03280444608794318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,2048,0.01923022170861562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,2560,0.04787910977999369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,1024,0.020326221982638042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,1536,0.02148088812828064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,1536,0.016367110941145156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,2048,0.042542222473356456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,768,0.015119110544522604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,1024,0.013412444127930535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,1024,0.01590577761332194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,4096,0.07233599821726482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,512,0.010671111444632212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,768,0.014141332772043018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,768,0.01203644441233741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,256,0.00654755574133661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,1024,0.02832711074087355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,768,0.024449777272012498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,512,0.01108000013563368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,128,0.0035502223504914176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,256,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,512,0.024139554964171514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,2560,0.032511999209721885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,64,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,256,0.01850311126973894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,128,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,32,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,128,0.013759999639458127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,64,0.009033777647548253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,32,0.009305777649084726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,65536,0.32188354598151314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,1536,0.036815110180113055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,65536,0.1735057830810547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,512,0.011012444065676795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,51200,0.25010845396253795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,256,0.008238222036096785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,65536,0.15068533685472277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,16384,0.08985422054926555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,51200,0.13135111331939697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,128,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,12288,0.06980444325341119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,51200,0.12446399529774983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,16384,0.04940977692604065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,16384,0.047798222965664335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,10240,0.051727109485202365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,12288,0.03348711133003235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,12288,0.03648444347911411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,16384,0.05512444509400261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,8192,0.0425893333223131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,10240,0.031643556223975286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,12288,0.043135109874937266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,10240,0.03132889005872939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,7168,0.038055999411476985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,8192,0.02472177810139126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,10240,0.03828088773621453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,65536,0.17784533235761854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,6144,0.03305244445800781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,7168,0.03159644537501865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,8192,0.034549332327312894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,7168,0.02201955517133077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,5120,0.027465777264700994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,6144,0.02676533328162299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,6144,0.02014044423898061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,4096,0.023715555667877197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,5120,0.024682665864626568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,5120,0.017489777670966256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,7168,0.03138488862249587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,6144,0.029734220769670274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,3584,0.02051377793153127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,4096,0.020149333609475028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,4096,0.015512888630231222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,3072,0.01788444485929277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,3584,0.018505778577592637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,4096,0.022118222382333543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,3072,0.016513778103722464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,3584,0.020781333247820537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,2560,0.015138667490747241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,8192,0.027245332797368366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,3072,0.013184888495339287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,2048,0.012340444657537671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,51200,0.13840177324083117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,2560,0.012331555287043253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,3072,0.019553777244355943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,2560,0.01457155578666263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,1536,0.009582222335868413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,2048,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,2048,0.013180444637934366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,2560,0.017314665847354464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,1024,0.007287110719415877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,1536,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,2048,0.016537777251667447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,768,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,1024,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,1024,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,5120,0.02628355556064182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,512,0.004217777815130022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,1024,0.013074666261672974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,3584,0.014360888136757744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,768,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,768,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,256,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,512,0.006664888726340399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,512,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,128,0.0022328889204396144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,768,0.011396444506115384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,256,0.005719999886221356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,512,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,64,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,128,0.00674044465025266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,128,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,256,0.009656888743241629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,32,0.002193777718477779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,128,0.007436444362004598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,64,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,32,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,65536,0.14984711011250815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,1536,0.014599111345079211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,51200,0.19906044006347656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,65536,0.1405537790722317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,51200,0.11269599861568874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,16384,0.06576710939407349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,51200,0.10367021958033244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,16384,0.041617777612474226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,12288,0.05102133419778612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,16384,0.03770133190684848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,16384,0.046795556942621865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,1536,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,12288,0.03196622265709771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,12288,0.029595557186338637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,65536,0.14486132727728948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,51200,0.11343555980258518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,12288,0.036497778362698026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,8192,0.03606933355331421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,10240,0.027331555883089702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,10240,0.02789777848455641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,10240,0.03307644526163737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,8192,0.024125332633654278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,8192,0.021998221675554912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,7168,0.031502223677105375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,256,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,6144,0.02767555581198798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,8192,0.029545777373843726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,7168,0.0274311105410258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,6144,0.02387199964788225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,6144,0.018128888474570382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,5120,0.02420355545149909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,65536,0.2536631160312229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,10240,0.04336533281538221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,4096,0.020601777566803824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,6144,0.025962667332755193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,5120,0.01623288955953386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,5120,0.022058667408095464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,3584,0.018258665998776753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,5120,0.022590221630202398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,4096,0.014663111832406787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,4096,0.01834755473666721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,3072,0.01611822181277805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,3584,0.017277333471510146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,3584,0.013560888667901358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,4096,0.019925332731670804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,2560,0.011519999967681037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,3072,0.015114666687117683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,7168,0.019931554794311523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,2560,0.013295110729005603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,3072,0.017100445098347135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,2560,0.011124444504578909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,2048,0.009813333551088968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,7168,0.02757688860098521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,1536,0.008504889077610439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,2560,0.016480889585283067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,2048,0.009662222531106737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,2048,0.012248888611793518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,1024,0.00666844430896971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,1536,0.009468444519572789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,2048,0.014744889405038623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,1536,0.010038221875826517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,768,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,1024,0.007380444142553542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,1024,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,1536,0.014156444205178155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,512,0.0041884444653987885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,768,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,768,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,1024,0.011520889070298938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,256,0.0028524444335036804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,3072,0.012514666550689273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,512,0.0069057775868309875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,3584,0.018569777409235638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,128,0.0024604443460702896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,512,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,768,0.01000088867213991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,256,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,64,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,256,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,32,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,128,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,512,0.009354666703277165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,128,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,256,0.009007111191749573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,64,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,65536,0.2177333301968045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,32,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,51200,0.17202044857872856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,65536,0.14215821690029568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,16384,0.058541331026289195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,65536,0.1254097753100925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,51200,0.10836711194780137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,51200,0.09728978077570598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,12288,0.04539288745986091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,16384,0.0395431121190389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,16384,0.03549955619706048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,10240,0.03748355640305413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,12288,0.028780443800820246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,12288,0.030609776576360066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,16384,0.0432204438580407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,10240,0.026798221800062392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,8192,0.031303998496797346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,12288,0.035303112533357404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,51200,0.106349335776435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,65536,0.13308444288041857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,7168,0.02797244323624505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,8192,0.021407110823525324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,128,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,8192,0.023030221462249756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,6144,0.024697777297761705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,7168,0.022272888157102797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,7168,0.01963733302222358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,5120,0.021204445097181533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,8192,0.028198222319285076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,6144,0.020219556159443326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,6144,0.01720977822939555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,7168,0.025661332739724055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,4096,0.018188445104493033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,5120,0.017861333158281114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,6144,0.02307022280163235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,5120,0.021013332737816706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,4096,0.015573332707087198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,3584,0.01588622232278188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,10240,0.02459200057718489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,4096,0.019018666611777413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,3072,0.014136888914638095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,3584,0.01389955480893453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,3584,0.01276088919904497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,10240,0.031672000885009766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,3584,0.018242667118708294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,2560,0.011613333390818702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,3072,0.011724444727102915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,3072,0.01313688854376475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,2048,0.009637333452701569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,2560,0.010072000324726105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,3072,0.016809778081046212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,2560,0.012074666718641916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,1536,0.008049777812427944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,2048,0.010703999963071613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,2048,0.009436444275909001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,5120,0.015451555450757345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,2560,0.015997333659066092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,1024,0.006778666542636023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,1536,0.008192000289758047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,4096,0.013812444276279874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,1536,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,2048,0.014751111467679342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,768,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,1024,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,1024,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,512,0.00388711111413108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,768,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,768,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,1024,0.010724444356229572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,256,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,512,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,512,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,768,0.010279999838935005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,128,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,256,0.0064151109092765385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,256,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,512,0.010057777994208867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,64,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,32,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,128,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,128,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,256,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,64,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,128,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,32,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,65536,0.09258844455083211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,51200,0.12966044743855795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,65536,0.06973333491219415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,51200,0.0692888895670573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,16384,0.04448089003562927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,1536,0.013904889424641928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,51200,0.05822933382458157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,12288,0.03459644317626953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,16384,0.028000000450346205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,16384,0.02276444435119629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,10240,0.026532444689008925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,12288,0.021558221843507554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,65536,0.1312719980875651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,12288,0.01848355597919888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,16384,0.04256177941958109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,8192,0.0233608881632487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,51200,0.10486577616797553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,12288,0.03492088781462775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,10240,0.01882755590809716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,7168,0.01960000064637926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,65536,0.17296266555786133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,8192,0.016140444411171805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,6144,0.017321777012613084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,10240,0.03095910946528117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,7168,0.013421333498424955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,8192,0.02664799988269806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,5120,0.014934221903483072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,7168,0.024247111545668706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,6144,0.02235200007756551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,5120,0.01423377792040507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,4096,0.012875555290116204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,5120,0.012039999994966718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,5120,0.020317332612143624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,4096,0.012366222010718452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,4096,0.010943999720944298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,3584,0.011113777756690979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,10240,0.017679111825095285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,8192,0.01480977733929952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,4096,0.01829155617290073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,3072,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,3584,0.009633777870072259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,3584,0.011409777734014722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,2560,0.008660444782839881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,7168,0.01789866718981001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,3072,0.008968888885445064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,3072,0.011174221833546957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,3584,0.01753155555990007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,6144,0.016480889585283067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,2048,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,2560,0.009963555468453301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,2560,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,3072,0.016533333394262526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,1536,0.006421333385838403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,2048,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,2560,0.015810666812790763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,2048,0.009031111167536842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,1024,0.004804444395833545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,1536,0.007358222372002072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,1536,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,768,0.0038942222793896994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,1024,0.006599111275540457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,1536,0.01204177737236023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,1024,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,768,0.006211555666393704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,512,0.002883555574549569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,768,0.010830221904648675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,512,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,256,0.00250133333934678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,6144,0.012568000290128918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,512,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,512,0.010061333576838175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,128,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,256,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,256,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,128,0.00565155554148886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,128,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,2048,0.014341332846217685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,256,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,64,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,32,0.0021893332401911416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,1024,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,128,0.007360888852013483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,64,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,32,0.005397333453098933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,768,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,51200,0.11598132716284858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,65536,0.08205866813659668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,16384,0.04062311185730828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,65536,0.06634400288263957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,51200,0.06294311417473687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,51200,0.05355644557211134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,12288,0.03184266553984748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,16384,0.025411556164423626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,16384,0.021700445148679946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,10240,0.026919111609458923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,16384,0.042303999265034996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,12288,0.01756000022093455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,12288,0.01975200076897939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,51200,0.10509777731365627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,8192,0.023253333237436082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,10240,0.017305778132544625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,10240,0.015790222419632804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,12288,0.03436977664629618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,8192,0.015549333559142219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,65536,0.14913422531551784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,8192,0.014071111877759298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,10240,0.0306915541489919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,8192,0.02660088903374142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,7168,0.012841777669058906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,7168,0.01685333251953125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,5120,0.013614222407341003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,6144,0.01197866681549284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,6144,0.015172445111804538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,7168,0.02387911081314087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,4096,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,5120,0.011008000208271874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,5120,0.013735111388895245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,6144,0.022649778260125056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,3584,0.010021333065297868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,65536,0.1308871110280355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,4096,0.012362666428089142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,5120,0.02015999952952067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,3072,0.008964444200197855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,3584,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,3584,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,4096,0.018580444984965853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,7168,0.017840888765123155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,3072,0.008208889100286696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,3072,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,3584,0.017208000024159748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,6144,0.015815999772813585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,2048,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,2560,0.00777688870827357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,2560,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,3072,0.016192000773217943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,1536,0.006223111102978389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,2048,0.007083555890454187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,2048,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,1024,0.004561777744028303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,1536,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,1536,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,2048,0.014261333478821648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,768,0.00388711111413108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,1024,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,1536,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,1024,0.006292444550328785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,4096,0.009718221922715506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,512,0.002874666617976295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,768,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,768,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,1024,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,256,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,2560,0.007632888853549957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,512,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,768,0.009721777505344814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,128,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,256,0.0058764442801475525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,256,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,512,0.009676444033781687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,64,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,128,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,128,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,256,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,2560,0.015487111277050443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,64,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,65536,0.12126133177015518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,128,0.0075804442167282104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,32,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,51200,0.09622933467229207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,65536,0.07413155502743192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,65536,0.0635626647207472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,16384,0.03414933217896356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,51200,0.05710577964782715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,51200,0.05154222249984741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,512,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,12288,0.02732622292306688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,16384,0.023356444305843774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,16384,0.02073777715365092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,51200,0.10479999913109674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,10240,0.023004444109068975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,16384,0.04200444287723965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,12288,0.017255110873116385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,32,0.001869333287080129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,8192,0.020198222663667467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,12288,0.01816444430086348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,10240,0.01516711049609714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,10240,0.015831111205948722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,7168,0.017886221408843994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,8192,0.013663999736309052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,12288,0.034041778908835515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,6144,0.01573333309756385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,8192,0.014477334088749357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,10240,0.02975733412636651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,7168,0.01276622215906779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,7168,0.016005333926942613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,5120,0.01402311192618476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,8192,0.026183111800087824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,6144,0.011721777419249216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,4096,0.010496888723638324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,7168,0.02370400064521366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,5120,0.010320000350475311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,5120,0.013106666505336761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,6144,0.0218248897128635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,3584,0.009277332988050249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,4096,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,4096,0.011945777469211154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,5120,0.019612444771660697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,3072,0.008260444634490544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,3584,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,65536,0.13039555814531115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,3584,0.010786666638321347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,4096,0.017840888765123155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,3072,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,3072,0.010363555616802638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,3584,0.017080888152122498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,2048,0.006456888798210356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,3072,0.016168889072206285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,2560,0.009022222624884712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,2560,0.007352888584136963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,1536,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,2560,0.01478844384352366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,1536,0.007000889215204451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,1536,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,2048,0.013491555220550962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,1024,0.003935111065705617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,6144,0.014429334137174817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,768,0.0031795555518733132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,1536,0.011729777687125735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,1024,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,1024,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,512,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,768,0.006022222340106964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,768,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,1024,0.010621333287821876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,256,0.002247111043996281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,2560,0.00758755538198683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,512,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,512,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,128,0.0021520000365045336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,256,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,512,0.010029333333174387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,64,0.0018880000958840053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,128,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,256,0.009364444348547194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,32,0.0018844445132546956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,2048,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,64,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,65536,0.10622221893734402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,128,0.007620444728268518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,32,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,51200,0.08465333117379083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,65536,0.06126044193903605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,768,0.009871111147933537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,65536,0.1305733389324612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,16384,0.031007111072540283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,51200,0.04952889018588596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,51200,0.10495377911461724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,128,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,2048,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,12288,0.024912888805071514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,16384,0.020184889435768127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,16384,0.021313778228229944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,10240,0.021149333980348375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,12288,0.01681866745154063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,12288,0.016521778371598985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,65536,0.06756799750857882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,12288,0.03336266676584879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,8192,0.018582221534517076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,10240,0.02934488985273573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,51200,0.0514977773030599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,8192,0.01347555551264021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,7168,0.016232000456915963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,8192,0.013168888787428537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,8192,0.026008889079093933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,6144,0.01443199978934394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,7168,0.012379555238617791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,7168,0.023461333579487268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,6144,0.013454222016864352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,5120,0.012443555725945367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,6144,0.01146577795346578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,16384,0.04138311081462436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,4096,0.010018666585286459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,6144,0.02182666626241472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,5120,0.009727111293209923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,10240,0.014780445231331719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,10240,0.014920888675583733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,5120,0.012441777520709567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,3584,0.008608889248636033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,4096,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,4096,0.011353777514563667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,3072,0.007809778054555257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,3584,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,5120,0.01960088809331258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,2560,0.006903999795516332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,4096,0.01827111177974277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,3072,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,3584,0.01718222267097897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,3072,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,2048,0.005836444596449534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,2560,0.0087013335691558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,7168,0.015470221638679504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,1536,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,2048,0.007024889190991719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,2560,0.015060444672902426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,1536,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,2048,0.01311555587583118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,1536,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,1024,0.003519111209445529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,1024,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,768,0.0032115555885765287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,1536,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,1024,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,1024,0.01074399964676963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,768,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,512,0.0028364445186323593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,768,0.005636444522274866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,512,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,768,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,256,0.002321777865290642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,512,0.005537777725193236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,2560,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,128,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,512,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,2048,0.007256000406212277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,64,0.0018968888454967076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,128,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,32,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,128,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,256,0.008635555704434713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,65536,0.08531822098626031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,64,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,128,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,32,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,51200,0.06665955649481879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,65536,0.06100177764892578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,65536,0.059106667836507164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,16384,0.025378665990299646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,3584,0.01035733355416192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,51200,0.04760533240106371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,16384,0.01924088928434584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,16384,0.019475556082195707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,12288,0.019208888212839764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,51200,0.10447111394670273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,3072,0.015583111180199517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,256,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,16384,0.04092799954944187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,10240,0.017166222135225933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,12288,0.015416888727082146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,10240,0.01444355481200748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,12288,0.03257600135273404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,10240,0.013807111316257052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,8192,0.014535110857751636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,8192,0.01255022237698237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,10240,0.028806222809685603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,7168,0.012977778083748288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,8192,0.01274222218328052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,7168,0.011383111278216044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,51200,0.047148442930645414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,6144,0.011928888658682505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,7168,0.01423288881778717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,65536,0.1304604477352566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,6144,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,7168,0.023504000571038988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,5120,0.010669333239396414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,6144,0.012847111456924014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,4096,0.011536888778209686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,5120,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,6144,0.02125066684352027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,5120,0.012069332930776807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,3584,0.011120888921949597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,4096,0.010374222364690568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,4096,0.008299555215570662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,5120,0.019644444187482197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,3072,0.009640889035330879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,3584,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,3584,0.009704888694816166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,4096,0.017834666702482436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,2560,0.006247111078765657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,3072,0.007709333466158972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,3072,0.008626666333940294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,3584,0.01680444512102339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,8192,0.025087111526065405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,2048,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,2560,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,2560,0.006993778049945831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,3072,0.014745778507656522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,1536,0.004522666749027041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,2048,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,2048,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,1024,0.0033520000676314035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,1536,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,2560,0.014131555954615274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,768,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,1536,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,2048,0.012748444245921241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,1024,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,12288,0.01612711118327247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,512,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,1024,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,1536,0.011366222467687396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,256,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,768,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,768,0.005629333357016246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,1024,0.010683555569913654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,512,0.005642666584915585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,128,0.0021377777059872947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,512,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,256,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,768,0.009688888986905416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,256,0.005044444567627377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,32,0.0018604445374674266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,128,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,512,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,128,0.004954666727119022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,65536,0.07663555277718438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,64,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,32,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,128,0.007621333003044128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,51200,0.05951378080579969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,65536,0.05750311083263821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,65536,0.05839288896984524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,16384,0.022359111242824133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,51200,0.04690044456058078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,51200,0.0450799994998508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,12288,0.018026666508780587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,65536,0.13053066200680202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,16384,0.018591110905011494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,16384,0.0192577772670322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,10240,0.015575110912322998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,51200,0.10445777575174968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,12288,0.015857777661747403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,12288,0.014936000108718872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,16384,0.04034666551484002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,8192,0.013551111022631327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,10240,0.014120888378885059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,10240,0.013395555317401886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,64,0.0019235555082559586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,7168,0.012055111428101858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,8192,0.012096000214417776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,8192,0.012376888758606382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,10240,0.02911111050181919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,256,0.009338666995366415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,7168,0.013211555778980255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,7168,0.01089777797460556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,8192,0.02513155506716834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,5120,0.010242666635248397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,7168,0.023205333285861548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,6144,0.009716444545321995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,6144,0.012401777837011548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,4096,0.01124888906876246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,6144,0.02128355536195967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,5120,0.011423110961914062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,4096,0.010104888843165504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,4096,0.008280889027648503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,3584,0.009999111294746399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,5120,0.019596444235907663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,4096,0.01751733322938283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,3072,0.00907022257645925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,3584,0.008295111358165741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,12288,0.03290133343802558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,3584,0.016131555040677387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,3072,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,2560,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,3072,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,3072,0.014496000276671516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,2560,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,2048,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,2560,0.013787555197874704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,5120,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,1536,0.0038862224254343244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,2048,0.006493333313200209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,2048,0.007073777417341868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,2048,0.012112889024946423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,1024,0.003167111012670729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,1536,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,1536,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,3584,0.007831999825106727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,768,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,1024,0.006016888966162999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,1024,0.006103111224042044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,6144,0.01124622258875105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,768,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,1024,0.010681777364677854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,768,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,768,0.010076444182131026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,256,0.0023493334237072202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,512,0.0052266667286554975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,512,0.009992889232105678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,256,0.004966222163703707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,128,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,256,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,256,0.008963555925422246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,64,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,128,0.005695111221737332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,32,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,1536,0.011735999749766456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,65536,0.0614008903503418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,64,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,128,0.007150222029950883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,32,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,51200,0.049088888698154025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,512,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,65536,0.054939554797278516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,65536,0.05724889039993286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,16384,0.019492444064882066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,512,0.005737777799367905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,2560,0.0074106670088238185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,51200,0.045964443021350436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,51200,0.04337955514589945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,12288,0.016311999824311998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,16384,0.018811555372344125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,65536,0.12995466921064588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,16384,0.01756444407833947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,10240,0.013463111387358772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,12288,0.015691555208630033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,51200,0.10426044464111328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,8192,0.012072888513406118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,12288,0.014479110638300577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,10240,0.013821333646774292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,7168,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,10240,0.013055110971132914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,8192,0.011769777370823754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,8192,0.011776888536082374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,6144,0.010218666659461128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,10240,0.028748446040683325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,7168,0.010660444696744284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,7168,0.012489777472284106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,8192,0.025506666964954797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,5120,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,6144,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,7168,0.022996443841192458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,6144,0.011750222080283694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,4096,0.008328888979223039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,5120,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,5120,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,6144,0.02126933303144243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,4096,0.008259555531872643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,5120,0.01945066617594825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,4096,0.016976000534163583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,3584,0.007826666865083907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,3584,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,3072,0.008600888980759515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,16384,0.04013244311014811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,3584,0.01584533353646596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,12288,0.03278844555219015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,2560,0.005633777628342311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,3072,0.014787554740905762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,2560,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,2560,0.007258666886223688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,2048,0.004857777721352047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,2560,0.013447110851605734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,2048,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,1536,0.0038284444146686127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,4096,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,2048,0.01274400038851632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,1536,0.006186666587988536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,1536,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,1024,0.003160888950030009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,1536,0.011247999966144562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,3072,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,3072,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,1024,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,1024,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,1024,0.010175111393133799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,512,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,768,0.005979555762476391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,768,0.005437333136796951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,768,0.010552888943089379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,512,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,256,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,512,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,3584,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,2048,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,128,0.0019031111150979996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,256,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,512,0.009014222357008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,256,0.005032888717121548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,128,0.005630222045713001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,32,0.0018213333355055915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,256,0.009335111412737105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,128,0.007689778175618913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,65536,0.05667022201750013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,64,0.005143110950787862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,32,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,51200,0.043990221288469106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,65536,0.056114666991763644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,65536,0.05311555663744608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,16384,0.01789955629242791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,51200,0.041950222518708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,51200,0.04524622360865275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,65536,0.13007733556959364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,12288,0.014016888207859464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,16384,0.01886399918132358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,16384,0.017196445001496207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,51200,0.10393777820799087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,10240,0.012258666257063547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,12288,0.015444444285498725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,16384,0.03976800044377645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,12288,0.014127999544143677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,8192,0.010718222293588849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,128,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,10240,0.013289777768982781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,10240,0.01274311128589842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,12288,0.03250311149491204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,8192,0.011058666639857821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,8192,0.011381333072980246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,6144,0.009087111386987898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,768,0.0028275555620590844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,10240,0.02918755677011278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,7168,0.01036622209681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,8192,0.02498488956027561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,5120,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,6144,0.009514666265911525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,7168,0.023001778456899855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,4096,0.007358222372002072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,5120,0.008796444369686974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,5120,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,6144,0.021190222766664293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,3584,0.008637333081828224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,5120,0.01891644464598762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,4096,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,4096,0.008096000386608971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,3072,0.008267555799749162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,3584,0.007939555578761632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,3584,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,4096,0.016886222693655226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,2560,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,3072,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,3072,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,7168,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,3584,0.015491555134455362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,2048,0.004578666554556953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,2560,0.007000000112586551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,2560,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,3072,0.01440177857875824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,7168,0.011730666789743634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,2048,0.006781333436568578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,2048,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,1024,0.0032008888406885993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,2560,0.013448889056841532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,6144,0.01034755590889189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,1536,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,2048,0.012039999994966718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,768,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,1024,0.005972444597217772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,1024,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,1536,0.011747555600272285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,512,0.0025315555847353404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,1024,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,768,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,512,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,256,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,768,0.009834667046864828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,512,0.009674666656388177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,256,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,128,0.001979555520746443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,1536,0.0037413334680928122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,256,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,64,0.0018177777528762817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,128,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,32,0.0017866666118303935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,128,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,1536,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,65536,0.04897422260708279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,128,0.007679999702506595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,64,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,32,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,65536,0.05077422327465481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,65536,0.05553244551022848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,512,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,51200,0.04064711266093784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,16384,0.014506666196717156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,51200,0.044510222143597074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,16384,0.01832088828086853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,12288,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,16384,0.016750221451123554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,16384,0.03987288806173537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,12288,0.014003554979960123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,10240,0.010484444598356882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,12288,0.014358222484588623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,12288,0.03252089023590088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,51200,0.03538489010598924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,10240,0.012467555701732635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,8192,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,10240,0.013051555388503604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,768,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,7168,0.009000889129108852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,10240,0.02851466668976678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,8192,0.011050666371981302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,6144,0.010696888797812991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,8192,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,65536,0.12934666209750706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,7168,0.010660444696744284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,7168,0.010087110930018956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,5120,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,8192,0.024844444460339014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,6144,0.009328000247478485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,4096,0.009043555292818282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,6144,0.009002666506502364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,51200,0.10423466894361709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,5120,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,3584,0.008635555704434713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,4096,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,4096,0.007485333416197035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,5120,0.01824711097611321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,3072,0.005990222096443176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,3584,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,3584,0.007666666474607255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,4096,0.016470222009552848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,2560,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,3072,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,3584,0.015433778365453085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,3072,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,2048,0.004281777888536453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,2560,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,2560,0.007122666471534305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,3072,0.014278221461508008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,1536,0.0038515557017591265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,2048,0.006680000159475539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,2560,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,1536,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,1536,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,2048,0.012444444828563266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,1024,0.003128000017669466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,7168,0.02330844435426924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,768,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,5120,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,6144,0.020611556039916146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,1024,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,1536,0.011696000066068439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,512,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,1024,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,768,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,768,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,256,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,1024,0.010656000011497073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,768,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,128,0.0018782222436534034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,256,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,256,0.005070222334729301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,512,0.009029332962301042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,256,0.009342222577995723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,128,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,128,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,32,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,2048,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,65536,0.03487111131350199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,128,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,64,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,32,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,51200,0.027344889110989038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,65536,0.04896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,65536,0.0547271105978224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,51200,0.039108445247014366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,51200,0.04429955614937676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,65536,0.12973422474331325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,16384,0.012467555701732635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,512,0.005199111170238919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,12288,0.010310221877362993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,16384,0.01735822194152408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,51200,0.1034159991476271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,16384,0.01620533400111728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,10240,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,12288,0.01316088851955202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,64,0.0018159999615616268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,12288,0.013773333695199756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,16384,0.03974399964014689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,8192,0.007841777470376756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,10240,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,10240,0.011992000043392181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,7168,0.00812888890504837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,8192,0.01036266651418474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,8192,0.010694222317801582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,6144,0.007596444752481248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,10240,0.028463111983405218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,8192,0.0243493335114585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,7168,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,7168,0.010021333065297868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,5120,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,6144,0.009365333451165093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,6144,0.008981333010726506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,7168,0.021923555268181696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,4096,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,5120,0.008662222160233391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,6144,0.019952888290087383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,3584,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,4096,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,5120,0.017854221993022494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,4096,0.0074035558435651995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,3072,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,3584,0.007280888656775157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,12288,0.03218133250872294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,3072,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,2560,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,3584,0.015124445160230001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,3072,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,3072,0.014194667339324951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,2048,0.005533333453867171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,2560,0.007152888509962294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,2560,0.006555555595291986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,2560,0.013394667042626275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,512,0.005631999837027655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,1536,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,2048,0.005997333261701796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,5120,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,1024,0.003015111097031169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,1536,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,2048,0.012757333616415659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,1536,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,768,0.00281866660548581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,1024,0.006329777754015393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,1024,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,1536,0.011354666617181567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,4096,0.016522667474216886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,512,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,768,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,768,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,1024,0.010581333604123859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,256,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,512,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,512,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,128,0.0018666666001081467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,768,0.010240000155236986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,256,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,512,0.009330666727489894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,256,0.004956444518433677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,64,0.0017875555074877208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,32,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,128,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,128,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,256,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,65536,0.024854222933451336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,64,0.004961777892377642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,128,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,51200,0.02008888953261905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,65536,0.05435111125310262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,65536,0.047212445073657565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,16384,0.013101333545313941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,51200,0.043785777356889516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,51200,0.03801244497299194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,3584,0.007782222496138678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,65536,0.12883910867902967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,12288,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,16384,0.016028443972269695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,51200,0.10307200087441339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,16384,0.015847999188635085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,10240,0.009873777627944946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,12288,0.013106666505336761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,12288,0.01242844429281023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,16384,0.03964444332652622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,8192,0.009424000150627559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,10240,0.011356444822417365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,10240,0.011765333513418833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,12288,0.03209155466821458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,7168,0.008691555923885768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,2048,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,8192,0.010669333239396414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,8192,0.009938666390048133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,6144,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,10240,0.027691556347741023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,32,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,7168,0.010001777774757808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,5120,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,7168,0.009467555416954888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,8192,0.02366222275627984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,6144,0.009339555270142024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,4096,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,7168,0.021683555510309007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,5120,0.008628444539176093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,5120,0.008176888856622908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,6144,0.019613333874278598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,4096,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,4096,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,5120,0.018577777677112155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,3072,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,3584,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,3584,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,4096,0.01613333324591319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,2560,0.004861333303981357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,3072,0.007338666253619724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,3072,0.006907555378145642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,3584,0.015116444892353482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,2048,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,2560,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,3072,0.014136888914638095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,1536,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,2048,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,2048,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,2560,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,1024,0.003927111211750242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,6144,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,1536,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,1536,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,768,0.0034977777136696708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,1024,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,1024,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,1536,0.011928000383906893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,512,0.0028275555620590844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,768,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,768,0.005893333504597346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,1024,0.010664000279373594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,256,0.002420444455411699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,512,0.005640888793600931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,512,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,768,0.009649777577983009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,128,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,256,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,256,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,512,0.009299555586444007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,2560,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,64,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,32,0.002011555557449659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,128,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,128,0.004605333424276776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,256,0.008681777450773451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,64,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,32,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,2048,0.012111110819710625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,51200,0.01697688963678148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,65536,0.04650755392180549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,65536,0.05376977721850077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,3584,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,16384,0.01423377792040507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,51200,0.04310577776696947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,51200,0.037689778539869524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,65536,0.12887733512454563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,12288,0.010021333065297868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,16384,0.014789332946141561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,16384,0.015754666593339708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,10240,0.00850933293501536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,16384,0.03899111019240485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,12288,0.012094222009181976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,10240,0.011787555283970304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,10240,0.011075555450386472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,12288,0.03154044349988302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,8192,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,65536,0.019739554988013375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,7168,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,10240,0.02774488925933838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,8192,0.009673777553770278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,128,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,8192,0.02366488840844896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,7168,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,6144,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,7168,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,7168,0.021617778473430212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,6144,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,5120,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,6144,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,51200,0.1027875542640686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,4096,0.006224888894293044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,5120,0.008796444369686974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,6144,0.019923556182119582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,5120,0.008245333201355403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,3584,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,4096,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,5120,0.018234666850831773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,4096,0.007248889240953658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,3072,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,3584,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,3584,0.007753777835104201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,8192,0.010390222072601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,4096,0.01651822196112739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,2560,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,3072,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,3072,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,3584,0.015167999598715039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,2048,0.004221333397759331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,2560,0.007068444457319047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,3072,0.014446222119861178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,1536,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,2560,0.013055110971132914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,2048,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,2048,0.012723555167516073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,1536,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,1024,0.0035422220826148987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,1536,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,12288,0.013110222087966071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,768,0.00300444434914324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,1024,0.005969777703285217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,1536,0.010998222563001843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,1024,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,512,0.002704888789190186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,768,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,768,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,1024,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,256,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,512,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,512,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,768,0.010399111443095738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,128,0.00198399999903308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,256,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,512,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,2048,0.0067528887755341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,128,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,128,0.005311111195219888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,64,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,256,0.008905777500735389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,768,32,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,65536,0.02776355544726054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,128,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,64,0.005055111315515307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,51200,0.022286223040686712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,65536,0.04565511147181193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,65536,0.05128177669313219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,16384,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,51200,0.038754665189319186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,65536,0.12832178009880915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,16384,0.015805333852767944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,16384,0.014596444037225513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,51200,0.10230133268568252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,256,0.00462311092350218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,16384,0.03839022252294753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,10240,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,12288,0.012127111355463663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,12288,0.013059555656380124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,12288,0.030886222918828327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,10240,0.011762667033407422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,8192,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,32,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,10240,0.0274000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,7168,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,8192,0.009721777505344814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,8192,0.010687111152542962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,8192,0.023346667488416035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,51200,0.0369466675652398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,7168,0.010033778018421596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,7168,0.009336888790130615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,12288,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,7168,0.021594666772418555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,6144,0.00908622228437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,6144,0.008914666871229807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,2560,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,4096,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,5120,0.008809777597586313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,6144,0.019934222102165222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,5120,0.008264000217119852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,3584,0.005531555662552516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,4096,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,4096,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,5120,0.018219555417696636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,10240,0.011017777853541903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,3072,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,3584,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,3584,0.007576888634098901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,4096,0.01643288963370853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,2560,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,3072,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,3584,0.015116444892353482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,6144,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,2560,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,2048,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,2560,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,3072,0.014081777797804939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,1536,0.0035075553589397003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,2048,0.0069066666894488865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,5120,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,2560,0.01310400002532535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,2048,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,1536,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,1536,0.005727111051479976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,2048,0.01200711147652732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,768,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,1024,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,1024,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,1536,0.011392888923486074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,512,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,768,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,1024,0.01000355515215132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,768,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,256,0.002201777779393726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,512,0.005199999858935674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,3072,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,512,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,768,0.00997511131895913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,128,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,256,0.005423999908897612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,512,0.009720000127951304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,64,0.0018053334206342697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,128,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,128,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,32,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,65536,0.0165928883685006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,256,0.009300444689061906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,128,0.007249777515729268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,512,1024,0.0030497778207063675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,51200,0.014144889182514615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,64,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,32,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,65536,0.04656799965434604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,65536,0.04344444473584493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,16384,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,51200,0.03701511025428772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,51200,0.03437599870893691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,12288,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,65536,0.12831021679772273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,16384,0.015787555111779105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,16384,0.014456000592973499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,10240,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,51200,0.1028124425146315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,12288,0.01309511148267322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,12288,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,16384,0.03831733266512553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,8192,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,10240,0.011728888584507836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,256,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,10240,0.010997333460383944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,7168,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,12288,0.031143112315071955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,8192,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,6144,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,8192,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,10240,0.027585777971479628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,7168,0.010025777750545079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,5120,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,7168,0.00926399976015091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,8192,0.02330933345688714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,6144,0.009140444298585257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,4096,0.004518222063779831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,6144,0.008873778084913889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,7168,0.021586666504542034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,5120,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,3584,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,5120,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,6144,0.0196079992585712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,4096,0.008151999778217739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,3072,0.004572444491916233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,5120,0.01822933389080895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,3584,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,4096,0.0158595558669832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,2560,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,3072,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,3072,0.006735111276308696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,2048,0.004683555414279302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,3584,0.015142222245534262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,2560,0.007105777661005656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,3072,0.014069333672523499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,1536,0.0035386664999855887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,2560,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,2048,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,2048,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,2560,0.013075555364290873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,1024,0.002943111169669363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,1536,0.005879999862776862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,2048,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,1536,0.011011555790901184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,1024,0.00599911105301645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,768,0.0025395556456512874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,1024,0.0052986666560173035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,4096,0.007010666860474481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,1024,0.01036088913679123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,768,0.005208000126812193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,768,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,512,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,256,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,512,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,512,0.009332444104883406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,256,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,128,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,1536,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,256,0.00869511150651508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,128,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,64,0.0018035556293196147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,768,0.005633777628342311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,128,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,32,0.0018071112119489247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,3584,0.007490666376219855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,65536,0.011402666568756104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,64,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,32,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,65536,0.04248266749911838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,65536,0.12736888726552328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,256,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,51200,0.03429155548413595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,16384,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,51200,0.03704444567362467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,128,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,16384,0.014484445254007975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,12288,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,51200,0.1009697781668769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,16384,0.015829333000712924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,256,512,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,10240,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,12288,0.013073777986897362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,16384,0.03798844416936239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,12288,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,8192,0.0052879999081293745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,10240,0.011750222080283694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,10240,0.010787555740939246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,12288,0.03127377894189622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,51200,0.009718221922715506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,8192,0.009690666364298927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,8192,0.010408889088365767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,10240,0.027053332991070215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,8192,0.023530667026837666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,7168,0.009360888765917884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,6144,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,6144,0.009353777600659264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,7168,0.022247999906539917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,5120,0.004580444345871608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,5120,0.00869777798652649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,6144,0.019578667150603402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,5120,0.008283555507659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,5120,0.01773066653145684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,4096,0.007114667031500075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,3584,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,4096,0.00795022232664956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,65536,0.04586933387650383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,7168,0.005140444470776452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,3072,0.004604444321658877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,3584,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,3584,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,7168,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,2560,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,3072,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,3072,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,2560,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,2048,0.0042257776690853965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,2560,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,4096,0.004607999904288186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,2560,0.013023110727469126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,1536,0.0032053333189752367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,2048,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,2048,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,2048,0.01237066669596566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,1536,0.006567111031876669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,1024,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,1536,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,4096,0.01648622254530589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,768,0.002536888958679305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,1024,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,1536,0.011071110765139261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,6144,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,1024,0.00566133318675889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,512,0.0024986666523747975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,768,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,1024,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,3584,0.014856000741322836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,256,0.0020008888095617294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,512,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,512,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,768,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,3072,0.014119111829333834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,512,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,256,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,128,0.005666666560702854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,64,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,128,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,32,0.0018044445249769424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,128,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,65536,0.009311111436949836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,64,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,32,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,51200,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,16384,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,65536,0.041787554820378624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,12288,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,51200,0.03387022349569533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,768,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,16384,0.014504000544548035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,10240,0.004562666846646203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,8192,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,7168,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,128,128,0.0017804443422291016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,12288,0.012040889097584618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,6144,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,256,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,10240,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,8192,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,5120,0.004217777815130022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,4096,0.004232888834344016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,3584,0.004512000001139111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,3072,0.004252444538805219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,256,0.00905511114332411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,2560,0.004563555535342959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,6144,0.008360889222886827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,5120,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,4096,0.007300444775157505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,3584,0.007090667055712805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,1536,0.003504000190231535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,1024,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,2560,0.006739555547634761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,768,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,2048,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,1536,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,512,0.0021537778278191886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,256,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,1024,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,128,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,768,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,64,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,512,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,256,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,128,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,65536,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,64,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,32,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,16384,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,65536,0.04201689031389025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,12288,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,2048,0.00423555572827657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,51200,0.034120000070995755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,10240,0.00452622233165635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,8192,0.004187555362780889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,3072,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,7168,0.004512888689835866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,12288,0.01202933324707879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,10240,0.010789333118332757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,6144,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,5120,0.004849777867396672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,8192,0.0096124443742964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,4096,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,7168,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,3584,0.004515555583768421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,6144,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,3072,0.004238222208287981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,64,32,0.0017893332988023758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,2560,0.0047155556579430895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,4096,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,2048,0.0041946665280395085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,3584,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,1536,0.0032702222880389956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,51200,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,1024,0.002891555635465516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,3072,0.006672888994216919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,768,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,7168,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,512,0.002505777817633417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,2048,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,256,0.0021439999755885866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,1536,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,128,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,1024,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,64,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,768,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,16384,0.014381332529915703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,16,32,32,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,512,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,256,0.004597333156400257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,128,0.004592000196377437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,64,0.004588444613748127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,32,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,5120,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,16384,0.3222213321261936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,16384,0.16312889258066812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,51200,0.4887395434909397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,12288,0.22687554359436035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,2560,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,51200,0.9703306621975369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,16384,0.1360657744937473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,12288,0.10083911154005264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,10240,0.18818222151862252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,8192,0.15391733911302355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,10240,0.1035733355416192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,10240,0.08795999818378025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,7168,0.1327422195010715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,8192,0.07046844561894734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,51200,0.43871021270751953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,6144,0.11438222726186116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,7168,0.09707822402318318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,8192,0.11178489526112874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,5120,0.09615200095706516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,5120,0.04619466596179538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,6144,0.05447999967469109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,7168,0.06404178010092841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,6144,0.08335644669002956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,5120,0.07103200091256036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,3584,0.06855199734369914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,3072,0.05942755275302463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,12288,0.12197155422634548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,4096,0.038903110557132296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,3584,0.035029331843058266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,4096,0.06272177563773261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,3584,0.05115644468201531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,2560,0.05027022295527988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,2048,0.04105510976579454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,3072,0.031217777066760596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,2560,0.02718488872051239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,3072,0.04546844296985202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,2560,0.03902755512131585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,2048,0.022898667388492163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,2048,0.03432710965474447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,1024,0.02236088944805993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,768,0.017811555001470778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,1536,0.018935999936527677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,1024,0.015041778484980265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,512,0.012653333445390066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,768,0.01331377774477005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,1024,0.019588443968031142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,256,0.007230222225189209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,768,0.015806222955385845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,128,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,512,0.01242844429281023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,4096,0.07962666617499457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,512,0.012558221817016602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,64,0.004228444563017951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,32,0.004193777839342753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,128,0.010401777923107147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,128,0.00816888858874639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,64,0.01000355515215132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,32,0.010029333333174387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,1536,0.03162844313515557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,1536,0.025774222281244066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,65536,0.510699537065294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,51200,0.3913866678873698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,65536,1.0887253019544814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,256,0.010416888528399997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,51200,0.861461321512858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,256,0.01000355515215132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,16384,0.27781687842475045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,12288,0.20628000630272758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,10240,0.17336177825927734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,16384,0.13112621837192112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,65536,0.4455217785305447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,12288,0.0835253331396315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,12288,0.09808355569839478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,10240,0.08319377899169922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,16384,0.10996711254119873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,51200,0.31811732716030544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,10240,0.06944621933831109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,8192,0.057757331265343555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,8192,0.09319200118382771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,7168,0.12433421611785889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,6144,0.10567377673255073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,5120,0.0890995528962877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,7168,0.07697777615653144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,7168,0.052143110169304736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,4096,0.07164977656470405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,6144,0.0449848903550042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,6144,0.07878933350245158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,5120,0.03767022159364488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,3584,0.06301155355241564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,5120,0.05685422155592176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,4096,0.0329146683216095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,3072,0.05537244346406725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,3584,0.028872887293497723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,2560,0.046133332782321505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,3584,0.04244266615973579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,3072,0.026748443643252056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,2048,0.03720533185535007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,3072,0.03586755527390374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,2560,0.022520000735918682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,1536,0.028966221544477675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,2560,0.03154222170511881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,1024,0.02083555526203579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,2048,0.028488000233968098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,8192,0.13944621880849203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,1536,0.016401777664820354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,768,0.015135111080275642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,1024,0.013421333498424955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,512,0.010632889138327705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,1024,0.016501333978441026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,768,0.011805333197116852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,256,0.00654666663871871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,768,0.013750221994188098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,512,0.01072177787621816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,128,0.0032764443506797156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,512,0.010944888823562197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,256,0.009378666679064432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,64,0.0028284444577164124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,256,0.008256889051861232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,128,0.009469333622190688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,32,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,4096,0.05158488949139913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,128,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,64,0.008971555365456475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,32,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,65536,0.3221777809990777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,2048,0.02000799940692054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,16384,0.08930222193400066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,65536,0.17587021986643472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,65536,0.15033245086669922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,1536,0.02241155505180359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,51200,0.13140622774759927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,12288,0.06996888584560819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,16384,0.04378755556212532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,51200,0.1257422235276964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,16384,0.0477164453930325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,8192,0.042917334371142916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,12288,0.03335377905103896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,12288,0.036582221587498985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,7168,0.037090665764278836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,10240,0.029118223322762385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,10240,0.031622221072514854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,8192,0.024329778220918443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,8192,0.03523733218510946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,7168,0.022346667117542688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,6144,0.03312089045842489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,5120,0.02828622195455763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,6144,0.020322667227851022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,6144,0.027772444817754958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,4096,0.023153778579499986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,5120,0.01772533357143402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,7168,0.03130311104986403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,3584,0.020576889316240948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,5120,0.02425866656833225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,4096,0.015832889411184523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,51200,0.24951288435194227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,3072,0.017887999614079792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,4096,0.020662221643659804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,2560,0.0151235560576121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,3584,0.018145778112941317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,10240,0.05208266774813334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,3072,0.013140444126394061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,2560,0.01237511138121287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,3584,0.014276444911956787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,2560,0.01425244410832723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,2048,0.010855110983053843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,1024,0.007369777394665613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,2048,0.013737777868906656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,1536,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,768,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,1536,0.011336888704035016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,1024,0.00795199970404307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,512,0.004181333300140169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,1024,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,768,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,256,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,512,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,768,0.007360888852013483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,128,0.002190222135848469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,256,0.006607111129495833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,64,0.0021688888470331826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,256,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,32,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,128,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,64,0.006217777729034424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,32,0.006182222316662471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,3072,0.016014221641752455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,2048,0.012272888587580787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,65536,0.2523769007788764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,1536,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,16384,0.06570400132073297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,51200,0.19898933834499785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,65536,0.14969778060913086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,12288,0.05092089043723213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,51200,0.11282400290171306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,65536,0.12999733289082846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,16384,0.037662221325768366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,10240,0.04327288932270474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,51200,0.10735555489857991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,16384,0.04129955503675673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,12288,0.029755555921130713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,8192,0.03654222355948554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,12288,0.032059556908077665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,512,0.006673777682913675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,10240,0.02574311031235589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,7168,0.031472888257768415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,128,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,6144,0.027521777484152053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,8192,0.02648711038960351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,8192,0.03007377849684821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,7168,0.027335110637876723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,6144,0.01817244456874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,6144,0.024010666542583044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,4096,0.019626667102177937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,5120,0.01612444387541877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,3584,0.016501333978441026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,5120,0.02085688875781165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,4096,0.019312888383865356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,4096,0.014458666245142618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,3072,0.014721777704026965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,3584,0.013296888934241401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,2560,0.012242666549152799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,3584,0.016415999995337594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,3072,0.012314666476514606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,2048,0.010568888651000129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,2560,0.011298666397730509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,2560,0.013424889081054263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,1536,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,2048,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,2048,0.01236977759334776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,10240,0.02793688906563653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,1024,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,1536,0.008574222524960836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,7168,0.019885333047972787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,768,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,1024,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,512,0.0038542221817705366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,1536,0.010334221853150262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,768,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,256,0.0028773333049482773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,768,0.0070159998204973005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,512,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,128,0.002475555572244856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,512,0.006404444575309753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,256,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,64,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,256,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,128,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,32,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,128,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,64,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,65536,0.2175431119071113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,32,0.00609777785009808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,65536,0.14168266455332437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,51200,0.1725991169611613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,5120,0.024001777172088623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,16384,0.05855821900897556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,65536,0.1268168952729967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,51200,0.09675911400053237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,12288,0.04547022117508782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,10240,0.03733333283000522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,16384,0.03604711095492045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,12288,0.028142223755518597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,8192,0.031209776798884075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,12288,0.030565334690941706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,10240,0.024890666206677754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,10240,0.02674399978584713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,7168,0.027777777777777776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,8192,0.020815999971495736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,6144,0.024725332856178284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,7168,0.019137778215938143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,3072,0.014570666684044732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,7168,0.022498667240142822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,5120,0.021151110529899597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,6144,0.020241777102152508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,6144,0.017400888933075797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,4096,0.018072888255119324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,5120,0.015503999259736804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,3584,0.015087111128701104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,4096,0.01368711143732071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,5120,0.017637333936161466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,51200,0.10763821999231975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,3072,0.013518222504191928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,3584,0.013006221916940479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,1024,0.008023999631404877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,2560,0.011389333340856763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,3584,0.01455200049612257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,3072,0.011737777955002256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,16384,0.03933066791958279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,2048,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,1536,0.008091555701361762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,3072,0.012846222354306115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,2560,0.01036977767944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,2048,0.009024889104896123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,2560,0.01146311147345437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,1536,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,2048,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,768,0.004876444323195351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,1536,0.009109333157539368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,8192,0.026015111141734656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,1024,0.007600000335110559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,512,0.0038933331767717996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,1024,0.007332444190979004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,768,0.00676711110605134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,256,0.0028373334142896864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,768,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,512,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,128,0.002456888970401552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,256,0.006481777876615524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,512,0.006059555543793573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,64,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,32,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,128,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,256,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,128,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,4096,0.016172443826993305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,64,0.005889777921968036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,65536,0.17350488238864473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,32,0.005631999837027655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,16384,0.04443644483884176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,65536,0.08790310886171128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,65536,0.07026133272382948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,12288,0.03511910968356662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,1024,0.006648889018429651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,51200,0.05784710910585192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,16384,0.023196443915367126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,10240,0.02674755619631873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,16384,0.02734222180313534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,12288,0.018577777677112155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,8192,0.02414488957987891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,12288,0.0214720004134708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,7168,0.02087466584311591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,10240,0.016433777080641854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,10240,0.01891999940077464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,8192,0.014542222023010254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,6144,0.018227555685573153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,8192,0.01982133256064521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,7168,0.0136817776494556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,5120,0.01549600064754486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,7168,0.017916444275114272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,4096,0.01310488912794325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,6144,0.01615111033121745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,5120,0.011672000090281168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,3584,0.011520889070298938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,5120,0.014766222900814481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,4096,0.010443555812040964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,3072,0.010327111515733931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,51200,0.13021600246429443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,3584,0.009667555491129557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,2560,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,3584,0.011740444435013665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,51200,0.06935466660393609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,3072,0.010804444551467896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,3072,0.00869777798652649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,2560,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,1536,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,2048,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,2048,0.00922400007645289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,1024,0.004671999977694618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,1536,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,1536,0.0076044441925154785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,768,0.003863111138343811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,6144,0.012780444489585029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,1024,0.006355555521117316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,512,0.0029022221763928733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,1024,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,256,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,768,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,512,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,128,0.0021671110557185277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,768,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,512,0.005635555419656966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,256,0.0058746664888328975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,4096,0.013377777404255338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,64,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,256,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,128,0.005879999862776862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,128,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,64,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,2048,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,32,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,51200,0.11654222011566162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,65536,0.08505600028567845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,16384,0.04073777794837952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,65536,0.06605066855748494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,12288,0.03177511029773288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,51200,0.06283733579847547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,51200,0.05368711219893562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,16384,0.021568000316619873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,10240,0.026896889011065166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,16384,0.025599999560250178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,12288,0.018059555027219985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,12288,0.020048000746303134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,10240,0.01564088960488637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,10240,0.0173973325226042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,7168,0.01955911186006334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,8192,0.013839110732078552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,6144,0.016857778032620747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,8192,0.019546666079097323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,7168,0.012757333616415659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,32,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,5120,0.014755555325084262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,7168,0.01685244507259793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,6144,0.012101333174440594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,5120,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,6144,0.015563555889659457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,2560,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,3584,0.011051555474599203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,5120,0.013769778112570444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,4096,0.009815110928482479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,3072,0.00980711148844825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,3584,0.008784000244405534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,3584,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,2560,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,3072,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,3072,0.010279999838935005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,8192,0.023168888356950548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,2048,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,2560,0.007675555845101674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,1536,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,2560,0.009584888815879822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,2048,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,1024,0.004530666602982415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,1536,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,2048,0.008623999853928884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,1536,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,768,0.003506666670242945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,1024,0.006308444258239534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,512,0.003020444470975134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,1024,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,768,0.005969777703285217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,65536,0.14926221635606554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,768,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,512,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,128,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,512,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,256,0.005774222314357758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,64,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,256,0.005231110999981563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,128,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,32,0.0019822222077184254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,128,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,65536,0.12191022766960992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,64,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,32,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,51200,0.09610133038626777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,65536,0.06428444385528564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,65536,0.0744782222641839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,16384,0.0344106654326121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,12288,0.027199111051029626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,51200,0.057020445664723717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,51200,0.051799999343024365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,16384,0.02086044516828325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,10240,0.023356444305843774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,16384,0.023522666758961145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,12288,0.017160000072585214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,8192,0.02010222276051839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,12288,0.018183110488785636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,4096,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,10240,0.015213333898120456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,7168,0.017816889617178176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,6144,0.015834665960735746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,8192,0.013901333014170328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,8192,0.01815200017558204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,5120,0.013978666729397245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,4096,0.012379555238617791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,6144,0.011751111182901593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,7168,0.015812445018026564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,4096,0.010610666539933948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,6144,0.014760888285107084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,3584,0.009289777941173976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,5120,0.010461333725187512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,4096,0.009309333231714036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,5120,0.013038222160604266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,3072,0.009188444250159794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,3584,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,2560,0.008044444852405125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,3584,0.010985777609878115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,3072,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,2048,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,2560,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,2560,0.008922666311264038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,1536,0.005881777654091517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,2048,0.0069120000633928515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,2048,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,256,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,1024,0.004109333372778363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,1536,0.006568000134494569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,10240,0.016188444362746347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,768,0.0033937777496046494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,7168,0.012553777959611682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,1024,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,512,0.0028968888024489083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,1024,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,768,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,256,0.0023493334237072202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,768,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,512,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,128,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,512,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,256,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,64,0.0017884444031450485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,256,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,4096,0.012071111136012606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,32,0.0018968888454967076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,128,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,65536,0.10752889182832505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,64,0.005471111171775394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,32,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,51200,0.08421066734525894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,65536,0.0680968893898858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,65536,0.06129778093761868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,51200,0.04955377843644884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,1536,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,51200,0.0513626668188307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,12288,0.02489600082238515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,16384,0.021199999584092036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,16384,0.02008800043000115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,10240,0.021003555920388963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,12288,0.016528889536857605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,8192,0.018537777993414137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,12288,0.016512889001104567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,10240,0.014895111322402954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,7168,0.016177778442700703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,10240,0.014896889527638754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,8192,0.013177778157922955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,128,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,8192,0.017536888519922893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,6144,0.014264000786675347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,5120,0.012381333443853589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,7168,0.012249777714411417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,7168,0.015170666906568738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,6144,0.011113777756690979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,4096,0.010850666297806634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,6144,0.013472889032628802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,5120,0.009863110880057016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,3584,0.009544889132181803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,4096,0.008352888955010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,5120,0.012383999923865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,3072,0.008660444782839881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,16384,0.03084889054298401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,4096,0.011419555379284753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,3584,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,2560,0.0074133334888352295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,3584,0.0103502223889033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,3072,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,2048,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,2560,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,1536,0.005167110926575131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,3072,0.008952000074916417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,2048,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,1024,0.00388355553150177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,1536,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,2048,0.007600000335110559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,768,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,1024,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,1536,0.0069102222720781965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,512,0.002805333377586471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,768,0.005976888868543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,256,0.0022400000856982335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,512,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,768,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,3072,0.00816533300611708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,128,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,512,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,64,0.0018737777653667664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,256,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,32,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,128,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,256,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,65536,0.08547555738025242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,128,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,51200,0.0666524436738756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,64,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,32,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,65536,0.058895111083984375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,65536,0.0607244438595242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,12288,0.020296888218985665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,51200,0.04760977625846863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,51200,0.047135111358430654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,16384,0.01958755486541324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,10240,0.017810665898852877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,16384,0.019327999817000497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,2560,0.007618666523032718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,12288,0.01608444419172075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,8192,0.015462223026487561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,7168,0.012831110921170978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,12288,0.015407111909654407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,10240,0.013784888717863293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,8192,0.012768000364303589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,6144,0.011334222224023608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,8192,0.01698844465944502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,7168,0.011710222396585675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,5120,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,7168,0.014480888843536377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,6144,0.010584000084135266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,4096,0.008936000367005667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,5120,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,6144,0.012940444052219391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,3584,0.00795911086930169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,5120,0.011704000333944956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,4096,0.010590222146775989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,3072,0.007060444189442529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,3584,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,3584,0.009233777721722921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,2560,0.0063573333124319715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,3072,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,3072,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,2048,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,2560,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,2560,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,1536,0.004111111164093018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,10240,0.0144195556640625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,2048,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,1024,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,2048,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,1536,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,1024,0.006216889040337668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,768,0.002855111120475663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,1024,0.006077333456940121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,512,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,1024,0.006123555617200003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,768,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,256,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,4096,0.008285333712895712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,768,0.00553955551650789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,512,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,128,0.0021377777059872947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,512,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,64,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,256,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,16384,0.025399110383457605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,256,0.005161777966552311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,128,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,128,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,65536,0.07698044512006971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,64,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,51200,0.05929244226879544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,32,0.005044444567627377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,16384,0.02236355510022905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,65536,0.058028446303473584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,65536,0.05800266398323906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,12288,0.017934223016103108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,51200,0.046774221791161426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,51200,0.045240001546012036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,10240,0.015490666031837463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,16384,0.01829688913292355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,12288,0.01591822173860338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,8192,0.013487999637921652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,1536,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,7168,0.012088889049159156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,12288,0.014885332849290637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,10240,0.014186667071448432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,6144,0.010960000256697336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,10240,0.013461333182122974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,8192,0.012135999898115793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,5120,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,7168,0.011078221930397881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,8192,0.015466666883892484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,4096,0.008647110727098253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,7168,0.013759111364682516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,6144,0.011682666838169098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,3584,0.007466666400432587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,5120,0.01146755533085929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,32,0.0019004444281260173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,3072,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,4096,0.009949333137936061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,3584,0.008202667037645975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,2560,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,3584,0.007932444413503012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,3072,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,2048,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,3072,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,2560,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,1536,0.003887999802827835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,2560,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,2048,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,1024,0.0031679999083280563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,2048,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,1536,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,768,0.0028826666788922418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,1536,0.006055111272467508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,1024,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,512,0.0024684444069862366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,1024,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,768,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,256,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,6144,0.01036711119943195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,768,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,512,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,128,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,5120,0.008842666943868002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,512,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,64,0.001800000046690305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,4096,0.008360000120268928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,32,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,128,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,65536,0.06170044342676798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,256,0.005151111218664381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,128,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,51200,0.04930577675501505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,64,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,32,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,16384,0.01942133406798045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,65536,0.057320886188083224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,65536,0.05580088827345106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,12288,0.015603555573357476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,51200,0.04601600103908115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,51200,0.04344177908367581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,16384,0.018894222047593858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,16384,0.017666666044129264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,12288,0.015658666690190632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,8192,0.01348000019788742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,10240,0.013909333282046847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,12288,0.014522666732470194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,10240,0.013076444466908773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,8192,0.011855999628702799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,6144,0.0110880004035102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,8192,0.013458666702111563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,7168,0.010737777584128909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,5120,0.009089777866999308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,7168,0.012871999707486896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,6144,0.009462222456932068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,4096,0.008248888783984715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,6144,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,5120,0.008752000000741746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,5120,0.010138666464222802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,4096,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,3584,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,3072,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,4096,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,3584,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,2560,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,3584,0.008014221986134848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,3072,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,2048,0.004555555681387584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,10240,0.015142222245534262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,3072,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,16384,0.019229332605997723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,1536,0.0038204445607132385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,2560,0.007199111084143321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,1024,0.0031724443866146933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,2560,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,2048,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,768,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,1536,0.006496888895829518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,2048,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,7168,0.012146666646003723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,1024,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,1536,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,256,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,768,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,128,0.0019022222194406723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,1024,0.005669333454635408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,768,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,512,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,64,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,256,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,512,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,128,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,65536,0.05655733082029554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,256,0.005052444421582752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,51200,0.04378755556212532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,64,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,32,0.00498577786816491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,16384,0.019659555620617337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,65536,0.05635199944178263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,65536,0.05338044299019707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,12288,0.01649777756796943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,51200,0.04522844486766391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,51200,0.04184177849027845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,10240,0.014135110709402295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,16384,0.018575110369258456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,12288,0.014987554815080432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,8192,0.012792888614866467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,16384,0.01724444495307075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,10240,0.01311911145846049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,12288,0.014118222726715935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,10240,0.012774222426944308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,8192,0.011026666396194033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,6144,0.01016177816523446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,7168,0.010060444474220276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,8192,0.012978666358523898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,5120,0.008817777865462834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,7168,0.01146755533085929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,6144,0.009365333451165093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,512,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,4096,0.00795022232664956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,3584,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,6144,0.010772444307804108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,5120,0.0084915558497111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,4096,0.008045333127180735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,3072,0.006221333311663733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,3584,0.00795911086930169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,4096,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,2560,0.0052266667286554975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,3584,0.007245333658324347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,3072,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,128,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,2048,0.004238222208287981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,2560,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,3072,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,1536,0.003656889001528422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,2560,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,2048,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,1024,0.0031715554909573663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,2048,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,768,0.0028799999919202593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,1536,0.006568888823191325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,1024,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,512,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,1536,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,7168,0.011406222151385413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,1024,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,256,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,768,0.005956444475385878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,128,0.0019324444648292328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,768,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,512,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,256,0.005640000104904175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,64,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,512,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,32,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,128,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,256,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,65536,0.04866666595141093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,128,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,5120,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,51200,0.03557777735922072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,16384,0.016150222884284127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,64,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,32,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,65536,0.055388443999820285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,12288,0.013214222258991666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,65536,0.05057155423694187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,10240,0.011962666279739805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,51200,0.04047466648949517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,12288,0.0136808885468377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,8192,0.011243555280897351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,16384,0.016777777009540133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,10240,0.012178666889667511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,7168,0.009666666388511658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,10240,0.012471111284361945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,8192,0.011008000208271874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,6144,0.008992888861232335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,8192,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,7168,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,5120,0.00832266691658232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,7168,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,6144,0.009315555294354757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,4096,0.007234666910436418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,6144,0.00907022257645925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,5120,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,3584,0.00656444455186526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,5120,0.008360889222886827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,4096,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,3072,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,4096,0.007631111476156447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,3584,0.007663999994595845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,2560,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,3584,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,51200,0.04455822375085619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,3072,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,2048,0.0042462220622433555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,16384,0.018437332577175565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,2560,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,3072,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,2048,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,1024,0.0032035555276605818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,12288,0.013906665974193148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,2048,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,1536,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,768,0.0028488888508743713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,512,0.0025048889219760895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,1536,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,768,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,1024,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,256,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,512,0.005872000008821487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,768,0.005416888743638992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,128,0.0019528888579871918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,512,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,256,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,64,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,32,0.0018097776919603348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,128,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,256,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,65536,0.03487199876043532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,128,0.004846222284767362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,64,0.005056888692908817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,32,0.005051555732885997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,65536,0.05470044414202372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,65536,0.04900000161594815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,16384,0.012710221939616732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,51200,0.044192890326182045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,51200,0.03899733225504557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,16384,0.017881777551439073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,2560,0.00665866666369968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,16384,0.016347555650605094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,10240,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,12288,0.013211555778980255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,1024,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,12288,0.013653332988421122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,8192,0.008954666554927826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,10240,0.011796444654464722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,10240,0.011759111450778114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,7168,0.008584889272848764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,8192,0.01069155583779017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,6144,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,8192,0.009959110783206092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,7168,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,5120,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,7168,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,6144,0.009672000176376766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,4096,0.006899555524190267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,51200,0.027652444110976324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,6144,0.008691555923885768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,5120,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,5120,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,4096,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,3072,0.005537777725193236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,4096,0.007596444752481248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,3584,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,2560,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,3584,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,3072,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,2048,0.004201777693298128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,2560,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,3072,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,1536,0.003674666500753827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,2048,0.006791111081838608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,2048,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,1024,0.003149333306484752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,1536,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,1024,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,1024,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,3584,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,768,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,768,0.005439110928111606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,1536,0.0038480001191298165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,512,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,12288,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,512,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,128,0.0019004444281260173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,512,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,256,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,2560,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,64,0.0018053334206342697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,32,0.001800000046690305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,128,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,128,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,65536,0.027633777923054163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,1536,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,64,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,32,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,65536,0.0544071098168691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,16384,0.010501333408885531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,65536,0.046912001238928906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,768,0.002798222212327851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,51200,0.043819556633631386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,12288,0.008420444197124904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,51200,0.03773422373665704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,10240,0.007744000189834171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,16384,0.0158933334880405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,16384,0.015389333168665567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,12288,0.013155555559529198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,8192,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,12288,0.012503999802801343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,256,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,10240,0.011117333339320289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,8192,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,6144,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,8192,0.009949333137936061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,7168,0.009717333647939894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,5120,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,7168,0.009387555221716562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,256,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,6144,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,6144,0.008866666919655269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,5120,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,3584,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,5120,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,51200,0.021985777550273474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,3072,0.0053671110007498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,4096,0.007424888511498769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,3584,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,2560,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,3584,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,3072,0.007245333658324347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,2048,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,2560,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,2560,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,1536,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,10240,0.011866666376590729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,2048,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,7168,0.0069848886794514125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,2048,0.006216889040337668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,1024,0.0038604442444112566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,1536,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,768,0.0032088889016045462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,1024,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,1536,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,512,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,1024,0.005532444351249271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,4096,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,768,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,256,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,768,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,128,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,512,0.00553955551650789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,512,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,256,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,4096,0.008160889148712158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,64,0.0018862223045693503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,32,0.001971555459830496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,256,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,65536,0.021913778450753953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,128,0.005547555370463266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,51200,0.01754577789041731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,64,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,32,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,3072,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,16384,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,65536,0.04622311062282986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,12288,0.00795199970404307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,51200,0.04345600141419304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,51200,0.03734933336575826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,16384,0.015624889069133334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,10240,0.0070186663005087115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,16384,0.01455200049612257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,12288,0.013195555243227216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,8192,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,12288,0.0121760004096561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,7168,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,10240,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,8192,0.0107351111041175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,6144,0.006079111248254776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,8192,0.009704888694816166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,7168,0.009961778091059791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,5120,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,7168,0.009360888765917884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,6144,0.009364444348547194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,4096,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,6144,0.008941333327028487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,5120,0.008640000389681922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,3584,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,4096,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,5120,0.008018666671382057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,128,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,3072,0.005511111269394557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,2560,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,3584,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,65536,0.05396177702479892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,2048,0.004218666503826777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,3072,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,3584,0.007167110840479533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,1536,0.004418666578001446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,3072,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,2560,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,2048,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,1024,0.003559999995761447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,2048,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,768,0.003173333282272021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,1536,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,1024,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,10240,0.011799111134476133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,512,0.00249422217408816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,1024,0.00564533347884814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,256,0.0021671110557185277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,768,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,512,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,768,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,128,0.002120888895458645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,512,0.005052444421582752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,64,0.0018542222678661346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,128,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,256,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,768,32,0.0017955555684036678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,65536,0.015808888607554965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,128,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,64,0.004850666556093427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,4096,0.0075902218619982404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,65536,0.05166133244832357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,32,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,51200,0.013295999831623502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,16384,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,65536,0.045496000183953174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,12288,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,2560,0.007227555745177799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,51200,0.03837600019243028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,16384,0.015841777125994366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,10240,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,51200,0.03709777858522203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,16384,0.014480888843536377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,1536,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,8192,0.005710222240951326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,12288,0.012053333222866058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,7168,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,10240,0.01108977778090371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,8192,0.010643555886215635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,6144,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,8192,0.00977955593003167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,7168,0.009865778187910715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,5120,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,7168,0.009266667068004608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,256,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,6144,0.009319999979601966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,4096,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,6144,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,3584,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,5120,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,4096,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,3072,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,3584,0.007664889097213745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,4096,0.007256000406212277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,2560,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,3584,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,3072,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,2048,0.004863111095296012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,3072,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,2560,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,1536,0.003656889001528422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,2560,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,12288,0.013094222380055321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,2048,0.006581333362393909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,1024,0.002862222285734283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,10240,0.011872000164455838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,768,0.002806222273243798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,1536,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,1536,0.006055111272467508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,1024,0.005984888722499211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,1024,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,768,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,256,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,768,0.005406222409672207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,512,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,512,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,5120,0.008022222254011366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,64,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,128,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,128,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,32,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,65536,0.010661332971519895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,64,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,32,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,51200,0.00944266633854972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,65536,0.04519022173351712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,65536,0.04646933409902784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,16384,0.005321777943107817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,51200,0.03717777795261807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,512,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,51200,0.03510933452182346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,12288,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,10240,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,16384,0.014548444085650973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,12288,0.013155555559529198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,8192,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,512,128,0.0018728888697094387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,10240,0.011756444142924415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,7168,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,10240,0.011055111057228513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,256,0.004872888740566042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,8192,0.010448888772063784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,6144,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,8192,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,7168,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,5120,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,7168,0.00923466682434082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,6144,0.009200889203283522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,4096,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,5120,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,6144,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,3584,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,4096,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,3072,0.004584889031118817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,4096,0.007242666350470648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,2048,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,3584,0.007034666836261749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,3072,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,3072,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,2048,0.004598222259018156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,16384,0.015795555379655626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,2560,0.0069208890199661255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,2560,0.006298666612969504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,1536,0.0035075553589397003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,2048,0.006864000111818314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,2048,0.005984000033802456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,1024,0.0028408887899584244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,1536,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,768,0.002495999965402815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,1536,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,1024,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,512,0.0024844445288181305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,1024,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,768,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,768,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,256,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,512,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,512,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,5120,0.007865777446164025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,128,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,256,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,64,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,128,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,2560,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,256,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,256,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,65536,0.00794044468137953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,128,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,51200,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,64,0.004578666554556953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,65536,0.045864890019098915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,32,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,16384,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,65536,0.04195644458134969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,12288,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,12288,0.012321777641773224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,51200,0.0372995568646325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,10240,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,16384,0.01551822159025404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,16384,0.014427555931939019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,12288,0.013080889152155982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,8192,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,10240,0.011721777419249216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,7168,0.00517866677708096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,10240,0.01072355525361167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,8192,0.010666666759385003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,6144,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,8192,0.010038221875826517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,7168,0.009851555857393477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,5120,0.004582222137186262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,7168,0.009125333693292405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,3584,0.007709333466158972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,6144,0.008278222547637092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,5120,0.008657777474986183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,5120,0.008051555189821456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,3584,0.004643555730581284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,4096,0.007703111403518253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,4096,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,3072,0.004587555511130227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,3584,0.007782222496138678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,3584,0.006792000184456508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,2560,0.004580444345871608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,3072,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,3072,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,2048,0.004592000196377437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,12288,0.012051555845472546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,2560,0.0069164443347189165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,1536,0.003543111185232798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,2048,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,2048,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,1024,0.002826666666401757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,1536,0.006310222049554189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,6144,0.009311999711725447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,1536,0.005674666828579373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,768,0.0026764445420768526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,4096,0.004550222307443619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,512,0.002504000026318762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,768,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,256,0.0021484444538752237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,768,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,512,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,512,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,128,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,51200,0.03410222133000692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,256,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,256,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,64,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,128,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,128,32,0.0018053334206342697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,65536,0.007596444752481248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,51200,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,128,0.004948444250557157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,16384,0.005372444374693765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,64,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,2560,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,12288,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,65536,0.04226311047871908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,10240,0.005450666778617435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,51200,0.033597333563698664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,8192,0.004859555512666702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,7168,0.004580444345871608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,16384,0.014496000276671516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,6144,0.004580444345871608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,12288,0.011968000067604912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,5120,0.004559999952713649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,10240,0.011037333144081963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,8192,0.00965244405799442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,4096,0.0042408886882993905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,1024,0.0058746664888328975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,3584,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,7168,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,3072,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,6144,0.00870400004916721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,1024,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,2560,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,2048,0.004527111020353106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,5120,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,1536,0.003264888914095031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,4096,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,3584,0.007009777757856581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,1024,0.0028799999919202593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,768,0.002540444541308615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,3072,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,2560,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,512,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,2048,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,256,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,128,0.0018311111877361934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,1536,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,1024,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,768,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,512,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,32,0.0018115554832749898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,65536,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,256,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,51200,0.007605333295133378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,128,0.004552000098758274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,16384,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,32,0.004961777892377642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,12288,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,65536,0.04179733329349094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,51200,0.03422844409942627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,10240,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,8192,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,16384,0.014126222994592456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,7168,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,12288,0.012124444047609964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,6144,0.00462755560874939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,10240,0.011016888750924004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,5120,0.004260444392760594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,7168,0.00906755526860555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,4096,0.0041911109454101985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,6144,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,3584,0.004622222234805425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,5120,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,64,64,0.0016115555125806068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,3072,0.004536889079544279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,2560,0.004547555413511065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,4096,0.007082666787836287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,32,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,2048,0.0038942222793896994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,1536,0.00351555562681622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,64,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,1024,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,2560,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,768,0.00249422217408816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,2048,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,512,0.0022035555707083806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,1536,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,256,0.0021608887861172357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,1024,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,128,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,768,0.005509333478079901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,64,0.0018053334206342697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,512,0.00518844442235099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,15,32,32,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,8192,0.009663999908500248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,256,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,128,0.0047182221379545005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,64,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,32,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,16384,0.3240133391486274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,3584,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,3072,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,16384,0.16398044427235922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,51200,0.4892426596747504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,12288,0.22596356603834364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,12288,0.12178844875759548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,10240,0.1882106728023953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,16384,0.13644177383846706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,12288,0.12827732827928331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,10240,0.10309066375096639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,7168,0.13291554980807835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,8192,0.11328800519307454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,6144,0.11472622553507487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,51200,0.4197591145833333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,7168,0.06320266591178046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,7168,0.10027911265691121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,5120,0.09624533520804511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,6144,0.0551493333445655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,6144,0.08591555224524604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,4096,0.0809697773721483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,5120,0.04694044589996338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,5120,0.07072977887259589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,3584,0.0685066646999783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,8192,0.15399111641777888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,4096,0.039649777942233615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,3072,0.05924266576766968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,3584,0.034987555609809026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,3584,0.05313333206706577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,8192,0.0712897777557373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,3072,0.04912088976966011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,3072,0.030955556366178725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,2048,0.04084355632464091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,2560,0.027446221974160936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,2560,0.0432204438580407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,1536,0.03168088859981961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,2048,0.03426400158140395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,2048,0.023382221659024555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,1024,0.02235466738541921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,1536,0.01884977850649092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,1536,0.024485333098305598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,51200,0.9654382069905599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,768,0.017871111631393433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,1024,0.014865777558750577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,1024,0.01960444450378418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,512,0.012560000022252401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,4096,0.06163644128375583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,256,0.00721777809990777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,768,0.013459555804729462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,512,0.012777778009573618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,128,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,10240,0.08595644765430027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,256,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,768,0.01601777805222405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,64,0.00416355538699362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,32,0.0042239998777707415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,128,0.010465777582592435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,256,0.010037333601050908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,128,0.008350222474998897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,64,0.010016000105275048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,32,0.010226666927337646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,65536,0.5147875679863824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,2560,0.050107555256949536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,51200,0.3911164336734348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,51200,0.8429759873284234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,65536,1.0882355372111003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,16384,0.27569421132405597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,65536,0.4264257748921712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,16384,0.13127022319369847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,12288,0.20735111501481798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,12288,0.09806844260957505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,12288,0.10497244199117024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,10240,0.17281156116061738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,8192,0.13998044861687556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,10240,0.06960711214277479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,51200,0.3436737855275472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,10240,0.08331466383404203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,7168,0.12218311097886826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,512,0.012709333664841123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,6144,0.10571110910839504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,8192,0.05616177452935112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,5120,0.08919644355773926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,7168,0.08000177807278104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,7168,0.052136888106664024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,4096,0.07225688960817125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,6144,0.06947733296288384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,3584,0.0635662211312188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,5120,0.038667556312349104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,4096,0.04965688784917196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,3072,0.05489066574308607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,3584,0.04070311122470432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,16384,0.11016266875796848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,2560,0.045863999260796436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,3072,0.025959110922283594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,3072,0.038309332397249006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,2048,0.037554668055640325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,2560,0.02276533345381419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,8192,0.09224355883068508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,2560,0.03504533237881131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,1536,0.029091556866963703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,2048,0.01960266629854838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,1024,0.020628444022602506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,2048,0.02847555610868666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,1536,0.0166284441947937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,6144,0.04611733224656847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,768,0.015121777852376303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,5120,0.057146668434143066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,1536,0.021320889393488567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,1024,0.013497778111033969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,1024,0.01686844395266639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,256,0.006527110934257507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,768,0.014479999740918478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,512,0.011041777829329172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,128,0.003544888976547453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,4096,0.032178666856553816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,3584,0.02898933490117391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,64,0.0028044444819291434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,256,0.009542221824328104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,32,0.002854222224818336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,128,0.009400000174840292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,256,0.008359111017651027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,64,0.008856000171767341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,128,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,32,0.009034666750166152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,65536,0.3713208834330241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,65536,0.17791288428836396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,51200,0.2714586787753635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,16384,0.08922221925523545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,51200,0.13180532720353869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,512,0.01070755554570092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,65536,0.16639288266499838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,12288,0.0696471134821574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,768,0.011739555332395764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,16384,0.04987200101216634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,10240,0.05166844526926676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,512,0.011054221954610614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,16384,0.047373334566752114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,51200,0.12923910882737902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,12288,0.040844443771574236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,12288,0.03660355673895942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,8192,0.04398577743106418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,7168,0.03803822067048814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,6144,0.03295644455485874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,10240,0.02955377764172024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,8192,0.02422399984465705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,10240,0.03171733352873061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,7168,0.022005332840813532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,5120,0.027738667196697656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,8192,0.03484355409940084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,7168,0.030841777722040813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,4096,0.02394666605525547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,3584,0.020441777176327176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,6144,0.020245333512624104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,5120,0.017625777257813346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,4096,0.015167999598715039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,3072,0.017708443933063082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,5120,0.02328444520632426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,3584,0.0147733340660731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,2560,0.0151235560576121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,3584,0.01790488925245073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,3072,0.013357333011097379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,2048,0.01239822225438224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,3072,0.016959110895792644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,2560,0.012144889268610211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,1536,0.009777777724795872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,2560,0.015118221441904703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,2048,0.010738666686746808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,1024,0.007584888901975419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,2048,0.013434666726324292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,1536,0.009222221871217092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,768,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,1024,0.00813422186507119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,512,0.0041866666740841335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,1024,0.009308444129096137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,768,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,256,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,768,0.007991111112965478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,512,0.007116444408893585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,128,0.002190222135848469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,512,0.006798222247097228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,256,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,64,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,256,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,128,0.006611555400821898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,128,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,4096,0.02050755586889055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,64,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,32,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,65536,0.25359733899434406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,51200,0.19817154937320283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,16384,0.06541866726345487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,65536,0.14827822314368352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,65536,0.1372035609351264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,51200,0.11267999807993571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,1536,0.01071466671095954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,12288,0.0506586664252811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,16384,0.037541333172056414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,10240,0.043143110142813794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,51200,0.11088977919684516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,16384,0.04152266515625848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,8192,0.03587377733654446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,12288,0.03194488750563727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,10240,0.02602133320437537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,7168,0.031347556246651545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,10240,0.02780088782310486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,32,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,8192,0.031488001346588135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,6144,0.027522666586769953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,8192,0.022266666094462078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,5120,0.024080889092551336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,7168,0.01979377700222863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,6144,0.018041777941915724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,7168,0.026718222432666357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,6144,0.02830933199988471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,6144,0.024317334095637005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,3584,0.016580444243219163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,5120,0.02021511064635383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,4096,0.014454222387737699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,3072,0.014532445205582513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,4096,0.01889511115021176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,2560,0.01238933371173011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,3584,0.013125333521101209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,3584,0.0162773331006368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,2048,0.010399111443095738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,3072,0.015319999721315173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,2560,0.011052444577217102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,1536,0.008488000267081791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,2560,0.013911111487282647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,2048,0.00980000032318963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,1024,0.006737777756320105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,2048,0.012422222230169507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,1536,0.01034577770365609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,768,0.0053173332578606075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,1024,0.007479999628331926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,1024,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,768,0.0070053330726093715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,4096,0.019329778022236295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,512,0.0038808890514903595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,768,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,256,0.0028702221396896574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,5120,0.016103999482260812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,512,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,128,0.0026568888376156488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,512,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,256,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,64,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,128,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,128,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,32,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,12288,0.035057776504092746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,64,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,1536,0.008646222452322641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,65536,0.21735910574595133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,32,0.005979555762476391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,51200,0.17232088247934976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,16384,0.058408002058664955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,65536,0.1387946605682373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,65536,0.12745955255296496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,51200,0.09768533044391209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,51200,0.10753689209620158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,12288,0.04521422253714668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,10240,0.037139554818471275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,16384,0.039461334546407066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,16384,0.03968000080850389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,12288,0.031382222970326744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,8192,0.030827555391523574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,256,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,12288,0.03052177694108751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,10240,0.024251555403073628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,10240,0.02664088871743944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,8192,0.021321778496106465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,6144,0.02457422183619605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,8192,0.026202667090627883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,7168,0.019139554765489366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,5120,0.02120977805720435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,7168,0.022222222553359136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,6144,0.01718222267097897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,4096,0.01759288873937395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,6144,0.020979555116759405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,3584,0.015096000499195524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,4096,0.013750221994188098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,5120,0.017735110388861764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,3072,0.013766222529941134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,4096,0.016210666961140104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,3584,0.012811555630630918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,3584,0.013918222652541267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,2560,0.011364444262451597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,3072,0.011747555600272285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,2048,0.009744889206356471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,3072,0.013237333132161034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,2560,0.010312889185216691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,1536,0.008086221913496653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,2560,0.0120666664507654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,2048,0.00906844437122345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,1024,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,2048,0.010961777634090848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,1536,0.008007110820876228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,768,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,7168,0.027777777777777776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,1024,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,1536,0.008614222208658854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,512,0.0038373333712418876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,256,0.0028799999919202593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,1024,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,768,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,128,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,768,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,256,0.006310222049554189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,512,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,64,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,256,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,32,0.002218666589922375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,128,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,5120,0.015697777271270752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,65536,0.17293243938022188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,64,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,32,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,65536,0.08807555834452312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,51200,0.1293964385986328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,16384,0.04436888959672716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,51200,0.06926488876342773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,51200,0.058463997311062284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,12288,0.03493155704604255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,16384,0.024705777565638225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,10240,0.026713778575261433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,3072,0.012642666697502136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,16384,0.027482666903071936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,12288,0.0199306673473782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,8192,0.023987554841571387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,12288,0.021538666552967493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,7168,0.020792000823550753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,10240,0.017051556044154696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,512,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,8192,0.014849777022997538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,10240,0.018972444865438674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,6144,0.01818933255142636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,8192,0.020038222273190815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,7168,0.014075555735164218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,5120,0.015467555986510383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,7168,0.017969777186711628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,128,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,4096,0.013459555804729462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,6144,0.012723555167516073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,5120,0.011543110840850405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,6144,0.01588888963063558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,3584,0.01165422217713462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,3072,0.010393777655230628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,5120,0.01426666643884447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,4096,0.013314666847387949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,3584,0.009750222166379293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,2560,0.008992888861232335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,65536,0.07009333372116089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,3072,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,2048,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,3072,0.010942222343550788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,2560,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,1536,0.006533333410819371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,2560,0.010370666782061258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,2048,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,1024,0.0046302220887608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,2048,0.00868711123863856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,1536,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,768,0.0038382220599386427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,1536,0.0069857777820693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,1024,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,512,0.003026666740576426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,1024,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,768,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,256,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,768,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,128,0.0021928888228204516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,512,0.005663110978073544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,256,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,64,0.0018746666610240936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,256,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,32,0.0020826665891541377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,128,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,128,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,4096,0.010695999695195092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,64,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,32,0.005543111099137201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,3584,0.011374221907721626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,51200,0.1157955593532986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,16384,0.04050488935576545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,65536,0.06588711341222127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,65536,0.08314755227830675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,12288,0.03198933270242479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,16384,0.021591110361946955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,51200,0.06293066342671712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,10240,0.026748443643252056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,16384,0.025320000118679468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,12288,0.019952888290087383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,8192,0.02307111190425025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,10240,0.01566844516330295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,7168,0.019554666346973844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,512,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,10240,0.01734844512409634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,8192,0.014102222190962898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,6144,0.01702311138312022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,8192,0.02028622229894002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,7168,0.012791111237472959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,5120,0.014595554934607612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,7168,0.017380444539917838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,6144,0.01202933324707879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,4096,0.012981333666377597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,6144,0.01572533282968733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,3584,0.011029333704047732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,5120,0.011072888970375061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,65536,0.1489537821875678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,5120,0.013785777820481194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,4096,0.009681777821646797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,3072,0.009729777773221334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,3584,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,2560,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,4096,0.012334222594896952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,3072,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,2048,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,3584,0.011026666396194033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,1536,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,2560,0.007660444411966536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,51200,0.05373333228958977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,2048,0.007231999602582719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,2560,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,1024,0.004262222184075249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,1536,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,12288,0.01777777738041348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,1536,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,1024,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,512,0.0028862222615215513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,768,0.006029333505365584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,256,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,1024,0.006317333214812809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,768,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,512,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,128,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,512,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,256,0.005728000154097875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,64,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,256,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,32,0.0018542222678661346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,128,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,128,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,65536,0.1216213305791219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,64,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,32,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,51200,0.09585422277450562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,65536,0.07386310895284016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,3072,0.010854221880435944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,16384,0.03417333298259311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,65536,0.06367644336488512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,51200,0.05686577823426989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,12288,0.02718755602836609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,768,0.00350222239891688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,10240,0.02329066726896498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,51200,0.051423112551371254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,16384,0.020980444219377305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,16384,0.023356444305843774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,8192,0.01990133358372582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,12288,0.017166222135225933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,7168,0.017779555585649278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,12288,0.0185208883550432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,10240,0.01516711049609714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,8192,0.013458666702111563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,6144,0.015783111254374187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,8192,0.018264888061417472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,7168,0.012714666624863943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,5120,0.013815999858909182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,7168,0.016065778003798593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,6144,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,4096,0.010988444089889526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,6144,0.01510933372709486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,5120,0.010416000253624385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,3584,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,5120,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,3072,0.009119110802809397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,4096,0.011903111305501727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,3584,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,2560,0.008045333127180735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,3584,0.010593777729405297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,3072,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,2048,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,3072,0.010327999790509542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,2560,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,2048,0.008279999925030602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,1536,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,2560,0.009529777699046666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,1024,0.0041724443435668945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,2048,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,1536,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,2048,0.007664889097213745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,1024,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,1536,0.006646222124497096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,512,0.0028479999552170434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,10240,0.016027554869651794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,1024,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,256,0.002199999988079071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,768,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,512,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,128,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,256,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,512,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,64,0.0018728888697094387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,32,0.0018311111877361934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,256,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,128,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,4096,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,128,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,65536,0.10645066367255317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,64,0.005306666509972678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,51200,0.08404266834259033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,16384,0.030811554855770532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,65536,0.06128622425927056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,12288,0.02482755482196808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,51200,0.04927911029921638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,51200,0.05124977893299527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,10240,0.02093244426780277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,16384,0.02122666603989071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,12288,0.016568000117937725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,8192,0.01827911039193471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,768,0.0032133333798911837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,12288,0.016516443755891588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,10240,0.014987554815080432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,10240,0.014880888991885714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,8192,0.01314755529165268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,6144,0.014288889037238227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,8192,0.01817244456874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,7168,0.012278222375445895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,5120,0.012396444049146442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,7168,0.015051555302408008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,6144,0.011075555450386472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,6144,0.014783999986118741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,5120,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,5120,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,32,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,4096,0.011353777514563667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,3584,0.009569777382744683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,4096,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,3072,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,65536,0.06733066505855985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,16384,0.020263999700546265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,3584,0.008363555702898238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,2560,0.0075493330756823225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,3072,0.007703999678293864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,3584,0.00998488896422916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,2048,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,2560,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,3072,0.00926399976015091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,1536,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,7168,0.016061334146393668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,2560,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,1024,0.0036071112586392295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,2048,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,768,0.0031475555151700974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,1024,0.006265777680608962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,1024,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,512,0.0028497777465316984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,768,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,4096,0.011360000405046674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,768,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,256,0.0022968887868854734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,512,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,128,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,256,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,512,0.0053280000057485365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,64,0.001975999938117133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,256,0.005207111024194294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,32,0.0018168888572189545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,128,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,128,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,65536,0.0854568878809611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,64,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,32,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,51200,0.06656177838643391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,65536,0.05921866496404012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,2048,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,65536,0.061187558703952365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,16384,0.02532088922129737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,1536,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,51200,0.04773866799142626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,12288,0.02025422288311852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,51200,0.04708533485730489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,10240,0.017775999175177682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,16384,0.019549333386951022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,8192,0.015850666496488783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,12288,0.01604977746804555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,16384,0.018889778190188937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,10240,0.0144177774588267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,12288,0.015451555450757345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,7168,0.012884444660610624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,6144,0.011329777538776398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,8192,0.01276533305644989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,7168,0.011688888900809817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,8192,0.017108443710539076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,5120,0.010018666585286459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,7168,0.014121777481502958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,6144,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,4096,0.008647110727098253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,3584,0.007731555236710443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,6144,0.013768889009952545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,4096,0.008168000313970778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,5120,0.011689778003427716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,3072,0.007033777733643849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,4096,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,3584,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,2560,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,3584,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,2048,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,3072,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,2560,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,3072,0.008353778057628209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,1536,0.003915555361244413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,2048,0.006680888848172293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,2560,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,1536,0.006567111031876669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,1536,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,2048,0.0070186663005087115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,768,0.002862222285734283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,1536,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,1024,0.006060444646411472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,512,0.002500444443689452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,768,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,1024,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,256,0.002181333386235767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,768,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,128,0.002130666747689247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,256,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,512,0.005375111268626318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,64,0.0018035556293196147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,5120,0.009053332938088311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,256,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,128,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,32,0.0018977777411540349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,128,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,65536,0.07679555813471477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,64,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,51200,0.05922222137451172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,32,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,65536,0.05819288889567057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,65536,0.05807377894719442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,16384,0.02235466738541921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,1024,0.003220444545149803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,51200,0.04689955711364746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,12288,0.017882666654056974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,51200,0.045260445939169995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,16384,0.019299555155966017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,16384,0.018426666657129925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,12288,0.015835555063353646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,8192,0.013391999734772576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,10240,0.013760888742076026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,12288,0.014868444866604276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,512,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,10240,0.014208889669842191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,7168,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,10240,0.013437333206335703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,6144,0.010886222124099731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,8192,0.01237155579858356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,5120,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,7168,0.011412444214026133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,8192,0.015023999743991427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,4096,0.00867911097076204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,6144,0.010037333601050908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,7168,0.013194666968451606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,6144,0.01292266696691513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,3584,0.0075911109646161394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,5120,0.008953777452309927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,3072,0.006882666713661618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,5120,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,3584,0.007879111501905654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,4096,0.008903111020723978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,2560,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,3584,0.008030222521887885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,2048,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,3072,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,2560,0.00700444479783376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,1536,0.0038897775941424896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,2560,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,2048,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,10240,0.015468445089128284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,2048,0.006688000013430913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,1536,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,768,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,1536,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,1024,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,512,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,1024,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,768,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,256,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,768,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,512,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,128,0.0019048889064126545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,512,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,256,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,256,0.005056000004212062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,128,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,32,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,128,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,65536,0.06226133637958103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,3072,0.007600000335110559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,64,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,51200,0.04928266670968798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,32,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,16384,0.019274666905403137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,65536,0.057265778382619224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,65536,0.05553422371546427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,1024,0.003165333428316646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,12288,0.016155555844306946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,51200,0.04589777853753832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,51200,0.04352622230847677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,10240,0.015204444527626038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,16384,0.018920888503392536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,8192,0.01346577786737018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,16384,0.017535110314687092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,12288,0.015636444091796875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,7168,0.012089778151777057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,8192,0.011730666789743634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,12288,0.014395554860432943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,10240,0.013134222063753339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,6144,0.011057777537239922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,7168,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,8192,0.014441778262456259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,7168,0.01205599970287747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,4096,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,5120,0.008971555365456475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,6144,0.011758222348160215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,3584,0.007229333122571309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,5120,0.00980799976322386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,4096,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,3072,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,4096,0.008655110994974772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,3584,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,2560,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,3584,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,3072,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,3072,0.007141333487298753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,2048,0.004705777598751916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,2560,0.007221333682537079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,2560,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,1536,0.0038364442686239877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,2048,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,10240,0.013833777772055732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,2048,0.00656088896923595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,1024,0.003208000005947219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,1536,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,768,0.0028159999185138275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,1536,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,6144,0.009651555783218807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,512,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,1024,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,768,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,256,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,768,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,512,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,4096,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,128,0.0019600000232458115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,512,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,256,0.005630222045713001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,64,0.001863111017478837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,32,0.0018595556418100994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,256,0.005046222358942032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,128,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,65536,0.0566577778922187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,128,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,64,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,51200,0.043848001294665866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,32,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,16384,0.01983022193113963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,65536,0.05605600277582804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,65536,0.05338933401637607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,12288,0.016367110941145156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,51200,0.04174400038189358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,16384,0.018901333212852478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,10240,0.014069333672523499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,16384,0.0169840008020401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,5120,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,1024,0.005970666805903117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,8192,0.012803555362754397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,12288,0.015332443846596612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,7168,0.011304000185595619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,12288,0.014084445105658637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,10240,0.012894222305880653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,6144,0.010249777800507015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,10240,0.01277511070171992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,8192,0.013339555925793119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,5120,0.008711111214425828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,7168,0.011362666885058085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,6144,0.009687110781669617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,4096,0.007748444047239091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,6144,0.010553778045707278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,5120,0.008960000342792934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,3584,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,4096,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,4096,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,3584,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,3584,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,51200,0.04522844486766391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,2560,0.005395555661784277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,3072,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,3072,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,2048,0.004295999805132548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,2560,0.007185777856243982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,1536,0.0038942222793896994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,2048,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,2560,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,8192,0.011427555647161273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,1024,0.003165333428316646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,2048,0.006360000206364526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,768,0.00282577777074443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,1536,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,7168,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,512,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,1024,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,768,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,1024,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,256,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,512,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,128,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,512,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,256,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,64,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,128,0.005533333453867171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,128,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,64,0.005314666777849197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,32,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,65536,0.049944000111685864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,65536,0.055163555675082736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,51200,0.03552533189455668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,65536,0.05063733458518982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,1536,0.005970666805903117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,16384,0.016192000773217943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,768,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,51200,0.044702221949895225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,51200,0.040397332774268255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,256,0.005631999837027655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,12288,0.013280889226330651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,16384,0.01681688924630483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,16384,0.018191110756662156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,10240,0.011872000164455838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,3072,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,12288,0.014193778236707052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,32,0.0018746666610240936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,12288,0.013935999737845527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,10240,0.012075555821259817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,7168,0.009647111097971598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,10240,0.012489777472284106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,8192,0.011051555474599203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,8192,0.011709333293967776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,7168,0.009993777506881291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,5120,0.008243555823961893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,7168,0.010398222340477837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,6144,0.009576000273227692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,4096,0.007343110938866933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,6144,0.009015999734401703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,5120,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,3584,0.006519111080302133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,5120,0.008295999632941352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,4096,0.008060444560315874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,3072,0.005715555614895291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,4096,0.007463999920421177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,3584,0.007879999776681265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,2560,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,3584,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,5120,0.008540444076061249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,3072,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,2048,0.00425155543618732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,2560,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,1536,0.003899555653333664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,2560,0.006581333362393909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,8192,0.011277333729796939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,1024,0.0031946665710873077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,2048,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,1536,0.0063582224150498705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,768,0.002814222127199173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,6144,0.008961777720186446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,1536,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,1024,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,512,0.002510222295920054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,1024,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,768,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,512,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,768,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,128,0.0019217777169413036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,512,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,256,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,64,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,256,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,32,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,128,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,65536,0.03496177660094367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,64,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,32,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,51200,0.027613333529896204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,3072,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,65536,0.04896622233920627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,65536,0.055099556843439736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,16384,0.01274577776590983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,51200,0.04418844315740797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,16384,0.01717866626050737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,16384,0.016180444094869826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,256,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,10240,0.010001777774757808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,12288,0.013384000294738345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,12288,0.013515555196338229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,8192,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,10240,0.012063110868136087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,7168,0.008292444050312042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,8192,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,128,0.004941333499219683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,6144,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,7168,0.010041777458455827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,7168,0.009327111144860586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,5120,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,6144,0.009376000199053023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,6144,0.008915555146005418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,2048,0.006653333289755716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,4096,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,5120,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,5120,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,3584,0.005974222388532426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,12288,0.010994666980372535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,3072,0.005491555564933353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,4096,0.00757244477669398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,3584,0.00775822252035141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,2560,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,3584,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,3072,0.007359111474619971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,2048,0.004209777961174647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,3072,0.006907555378145642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,2560,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,1536,0.003555555724435382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,2560,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,10240,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,2048,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,1024,0.003141333245568805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,2048,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,1536,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,768,0.0028213332924577924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,1536,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,1024,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,512,0.0025519999778932994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,1024,0.005535999933878581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,256,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,512,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,512,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,128,0.0019075555933846368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,51200,0.03884177737765842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,256,0.005326222214433882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,64,0.0018195555441909367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,128,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,32,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,128,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,65536,0.028121779362360638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,64,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,32,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,8192,0.0107360002067354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,51200,0.021983999345037673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,65536,0.05420444409052531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,65536,0.04710488849216037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,16384,0.01036266651418474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,51200,0.04382488793796963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,12288,0.008618666893906063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,51200,0.03779733180999756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,16384,0.015823110938072205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,10240,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,16384,0.014857777290874057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,768,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,12288,0.01313066648112403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,8192,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,12288,0.01239466667175293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,10240,0.011784888803958893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,7168,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,4096,0.008172444171375698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,8192,0.01072177787621816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,10240,0.01128799964984258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,256,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,7168,0.010016000105275048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,6144,0.009320889082219865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,7168,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,4096,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,6144,0.008999110923873054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,5120,0.008645333349704742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,5120,0.008067555725574493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,3584,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,4096,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,3072,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,4096,0.0073520003093613525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,3584,0.007639110916190677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,2560,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,3584,0.007254222200976477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,3072,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,2048,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,3072,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,2560,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,1536,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,2560,0.0063271112740039825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,2048,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,1024,0.0038737778862317405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,2048,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,1536,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,768,0.0034719999465677473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,1024,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,1536,0.005980444451173146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,6144,0.007378666765160031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,512,0.0028080000645584534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,768,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,1024,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,256,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,5120,0.006528888725572162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,128,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,768,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,512,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,64,0.0018097776919603348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,256,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,32,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,128,0.0052968888647026485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,65536,0.021709332863489788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,256,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,128,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,51200,0.017684444785118103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,64,0.005045333256324132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,65536,0.05424444542990791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,32,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,16384,0.009284444153308868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,12288,0.007896000312434303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,65536,0.04628800021277534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,51200,0.037300444311565824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,16384,0.015840000576443143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,10240,0.0069288888739215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,16384,0.014783110883500842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,12288,0.012873777912722694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,8192,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,12288,0.012421333127551608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,10240,0.011718221836619906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,7168,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,10240,0.011077333655622272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,8192,0.010680889089902243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,6144,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,8192,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,7168,0.010021333065297868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,5120,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,8192,0.009718221922715506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,7168,0.009315555294354757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,6144,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,4096,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,6144,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,3584,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,5120,0.008635555704434713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,512,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,4096,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,3072,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,5120,0.008351999852392409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,2560,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,3584,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,4096,0.00739555557568868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,2048,0.0041697778635554844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,3072,0.0073626670572492815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,3584,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,51200,0.04301777813169691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,2560,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,3072,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,1024,0.0035093331502543557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,2560,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,2048,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,768,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,1536,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,2048,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,1024,0.006223111102978389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,512,0.002536888958679305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,1536,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,256,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,768,0.005657777604129579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,1024,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,128,0.0019662222928471034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,512,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,256,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,512,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,32,0.0018622221218215094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,256,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,128,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,65536,0.01575555569595761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,128,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,51200,0.013273778061072031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,64,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,65536,0.05112533436881172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,32,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,16384,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,65536,0.04591200086805555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,51200,0.038626667526033186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,12288,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,51200,0.03665511144532098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,16384,0.015746666325463187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,10240,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,768,1536,0.00453955555955569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,16384,0.014446222119861178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,8192,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,12288,0.01295111080010732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,10240,0.011701333026091257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,7168,0.0053377776510185665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,12288,0.012183999849690331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,6144,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,10240,0.011006222003036074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,8192,0.010447999669445885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,5120,0.0063075555695427795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,8192,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,7168,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,4096,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,6144,0.009344889058007134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,7168,0.009097778134875828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,768,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,5120,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,6144,0.008976000050703684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,3072,0.0053191110491752625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,4096,0.007723555796676212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,2560,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,3584,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,4096,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,3072,0.00739911115831799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,2048,0.004582222137186262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,3584,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,1536,0.0035075553589397003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,3072,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,2560,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,1024,0.003093333293994268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,2560,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,2048,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,768,0.00282577777074443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,1536,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,2048,0.005975111077229182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,512,0.002528888897763358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,1024,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,1536,0.005881777654091517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,768,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,1024,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,256,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,768,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,512,0.0058764442801475525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,128,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,512,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,256,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,64,0.0018115554832749898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,32,0.0018568889548381169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,256,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,65536,0.01071200023094813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,128,0.004871110949251386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,512,3584,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,64,0.004967111266321606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,51200,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,32,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,5120,0.00851022203763326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,65536,0.04639288783073425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,65536,0.042464001311196216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,12288,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,51200,0.03702044486999512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,51200,0.03479377759827508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,10240,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,16384,0.014496000276671516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,12288,0.012904889053768583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,8192,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,12288,0.012116444607575735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,10240,0.011883555187119378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,7168,0.005377777748637729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,10240,0.011061333119869232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,8192,0.010401777923107147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,6144,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,8192,0.00964355551534229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,7168,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,5120,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,7168,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,6144,0.009304888546466827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,128,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,4096,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,6144,0.00869955536392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,5120,0.00869599978129069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,5120,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,4096,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,3072,0.004541333350870344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,16384,0.005370666583379109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,4096,0.007419555551475949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,3584,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,2560,0.004544888933499654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,3584,0.0069982219073507525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,2048,0.004597333156400257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,3072,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,16384,0.015829333000712924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,3072,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,2560,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,2048,0.006545777950021956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,2560,0.0064479998416370815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,1024,0.002987555538614591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,1536,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,768,0.002536000063021978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,1536,0.005746666755941179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,1024,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,512,0.002512000087234709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,768,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,256,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,768,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,512,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,128,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,256,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,512,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,64,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,256,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,32,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,128,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,65536,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,64,0.004951111144489712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,32,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,1536,0.0035573335157500375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,51200,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,65536,0.04623288909594218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,65536,0.04208533300293816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,16384,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,2048,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,51200,0.037019554111692644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,12288,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,1024,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,16384,0.015801777442296345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,10240,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,12288,0.013058666553762225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,16384,0.014408000641398959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,8192,0.005281777845488654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,10240,0.011722666521867117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,7168,0.005037333402368757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,10240,0.0110871113008923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,8192,0.010623999767833285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,6144,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,256,3584,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,8192,0.009639110830095079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,7168,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,7168,0.009139555195967356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,6144,0.009147555463843876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,128,0.004534222185611725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,6144,0.00889866633547677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,4096,0.004247111164861255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,5120,0.008451555338170793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,3584,0.00462755560874939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,5120,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,4096,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,3072,0.004519999855094486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,4096,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,3584,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,2560,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,3584,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,3072,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,3072,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,2560,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,2560,0.006247111078765657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,12288,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,1536,0.003536000019974179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,2048,0.006646222124497096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,2048,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,1536,0.0063075555695427795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,1536,0.005888000130653381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,768,0.002513777878549364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,1024,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,768,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,768,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,512,0.0024684444069862366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,512,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,256,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,512,0.00499199993080563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,51200,0.03392622205946181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,128,0.001867555495765474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,256,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,256,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,64,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,128,0.00535377777285046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,32,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,1024,0.002887111157178879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,128,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,65536,0.007347555624114142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,51200,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,5120,0.004239999999602635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,64,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,1024,0.005881777654091517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,16384,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,12288,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,32,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,10240,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,8192,0.0046728890803125174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,51200,0.0336168905099233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,7168,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,16384,0.014155555102560254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,12288,0.012042666474978128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,6144,0.004600888739029567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,10240,0.010878221856223213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,5120,0.004520888957712385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,4096,0.003907555507289039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,8192,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,3584,0.004252444538805219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,3072,0.0041857775714662345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,6144,0.00850577818022834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,5120,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,2560,0.00452888881166776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,128,2048,0.004602666530344221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,2048,0.0042275554604000514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,4096,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,1536,0.0034959999223550162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,3584,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,1024,0.0028782222006056043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,3072,0.006593777901596493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,768,0.002540444541308615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,512,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,2560,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,2048,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,256,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,1536,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,128,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,64,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,1024,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,768,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,64,32,0.001777777862217691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,512,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,65536,0.008296888735559251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,51200,0.007280888656775157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,256,0.005043555465009478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,128,0.004963555683692296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,16384,0.0052888890107472735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,65536,0.04185511006249321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,64,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,12288,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,10240,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,32,0.004548444516128964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,8192,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,65536,0.041758222712410815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,51200,0.03365866674317254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,6144,0.004210666649871402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,16384,0.014492443866199918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,12288,0.012057777908113269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,5120,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,10240,0.011039999624093374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,4096,0.004202666795916027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,7168,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,3584,0.004259555704063839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,7168,0.008989333278603023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,3072,0.004179555508825514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,6144,0.008363555702898238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,5120,0.008034666379292807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,2560,0.0042133331298828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,4096,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,2048,0.003891555385457145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,1536,0.003264888914095031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,3584,0.006989333364698622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,1024,0.00286666676402092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,3072,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,768,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,2560,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,512,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,2048,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,256,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,1536,0.0058542220956749385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,128,0.0017706666969590718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,1024,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,64,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,768,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,32,0.0015004444867372513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,512,0.0054551110499435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,256,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,14,32,7168,0.004572444491916233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,128,0.004557333472702238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,64,0.004467555632193883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,32,0.004594666676388847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,8192,0.00944266633854972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,16384,0.3227599991692437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,12288,0.22622222370571562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,51200,0.48929776085747617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,51200,0.977193726433648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,12288,0.12194932831658258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,10240,0.1878053347269694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,12288,0.1031128896607293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,8192,0.15410133202870688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,10240,0.08687822024027507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,8192,0.11527021725972493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,51200,0.4443422158559163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,10240,0.1030648880534702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,7168,0.1327244440714518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,8192,0.06965688864390056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,16384,0.16347644064161512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,6144,0.11411910586886936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,7168,0.09918310907151963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,5120,0.09618666436937119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,16384,0.13620000415378147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,6144,0.05459555652406481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,7168,0.06651200188530816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,4096,0.07994133234024048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,6144,0.08557244141896565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,5120,0.04691733254326714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,5120,0.0745333300696479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,3584,0.06833599673377143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,4096,0.039008888933393694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,3072,0.05927822324964735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,4096,0.06359822220272489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,2560,0.04992799957593282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,3584,0.03496088915401035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,2048,0.041251556740866765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,3072,0.031214220656288996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,3584,0.050345778465270996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,1536,0.03186488813824124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,3072,0.045496000183953174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,2560,0.02762311034732395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,2560,0.03956444395913018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,768,0.017853332890404593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,2048,0.023021333747439917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,1024,0.015503111812803479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,1536,0.027040888865788777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,512,0.01258933295806249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,1024,0.0196106665664249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,768,0.013358222113715278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,256,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,768,0.01662400033738878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,512,0.012768000364303589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,128,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,512,0.012632889052232107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,64,0.004226666771703296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,128,0.010401777923107147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,32,0.004179555508825514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,128,0.007985778152942657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,64,0.010048000348938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,32,0.010020444790522257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,1024,0.022654222117529974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,1536,0.018897778458065458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,65536,0.5154799885219997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,256,0.010398222340477837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,65536,1.0875608656141493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,51200,0.3905128902859158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,16384,0.2793377770317925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,51200,0.8471653196546766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,12288,0.20622311698065865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,16384,0.11024355226092869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,65536,0.4502408769395616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,2048,0.03458666801452637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,16384,0.1306017769707574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,12288,0.09789511230256821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,10240,0.17291111416286895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,51200,0.3170479933420817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,8192,0.14038044876522487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,10240,0.08302755488289727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,10240,0.0697724421819051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,12288,0.0855413344171312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,256,0.009874666730562845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,7168,0.12183466222551133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,6144,0.10529955228169759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,8192,0.056531555122799344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,8192,0.10818222496244644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,5120,0.08856622378031413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,7168,0.05134488807784187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,4096,0.07211733526653714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,6144,0.07039378086725871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,6144,0.04530400037765503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,5120,0.03931377662552728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,5120,0.06103110975689358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,3584,0.06275733311971028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,3072,0.054789334535598755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,3584,0.029490666257010564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,4096,0.03328533305062188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,4096,0.053119109736548535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,2560,0.04599022203021579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,3072,0.025892444782786902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,2048,0.036906666225857206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,3072,0.042542222473356456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,3584,0.040088001224729754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,2560,0.022795556320084467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,1536,0.028988444142871436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,2560,0.03087200058831109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,1024,0.02060622142420875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,2048,0.023756444454193115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,2048,0.02997866604063246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,1536,0.019301333361201815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,1536,0.02235377828280131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,1024,0.013354666531085968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,512,0.010641777680979835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,768,0.012051555845472546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,1024,0.016414221790101793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,256,0.006321777900060018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,768,0.014300444059901767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,512,0.011183111204041375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,128,0.0034391110142072043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,64,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,512,0.011076444553004371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,32,0.002855111120475663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,256,0.009344889058007134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,128,0.009373333719041612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,256,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,128,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,7168,0.09451199902428521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,32,0.008985777695973715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,64,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,65536,0.3707466655307346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,65536,0.1724177731408013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,51200,0.27167020903693306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,16384,0.08930755323833889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,51200,0.13141332732306585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,65536,0.15881244341532388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,12288,0.0694720016585456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,768,0.01512888901763492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,16384,0.042303111818101674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,10240,0.05167555477884081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,8192,0.043507556120554604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,12288,0.036375999450683594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,16384,0.04764266808827718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,51200,0.1247475544611613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,10240,0.02922399838765462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,6144,0.03294310967127482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,8192,0.02402577797571818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,8192,0.035600887404547796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,7168,0.02205955485502879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,6144,0.019816888703240287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,6144,0.027971555789311726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,4096,0.023386667172114056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,5120,0.017631111873520743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,5120,0.024853333830833435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,3584,0.020434666011068556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,4096,0.015493333339691162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,4096,0.019906666543748643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,3072,0.017823111679818895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,3584,0.01789155602455139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,3584,0.014228444960382251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,2560,0.015078221758206686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,3072,0.013066666821638743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,3072,0.01754488878779941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,12288,0.03329244587156508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,2048,0.012410666379663678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,7168,0.03702133231692844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,2560,0.014416888356208801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,2560,0.012075555821259817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,1536,0.00980266680320104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,1024,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,2048,0.01256533298227522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,5120,0.02748888896571265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,1536,0.009293333523803288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,2048,0.013421333498424955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,768,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,7168,0.03091199861632453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,1024,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,1536,0.011398221883508893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,512,0.004115555435419083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,1024,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,768,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,512,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,768,0.007788444558779399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,128,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,512,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,64,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,128,0.006714666469229593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,32,0.002198222196764416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,128,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,64,0.005975111077229182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,32,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,65536,0.25270220968458385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,10240,0.03182399935192532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,51200,0.19814756181504992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,65536,0.15205332967970106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,16384,0.06511911418702868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,65536,0.1340008841620551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,256,0.002529777793420686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,12288,0.050952888197369046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,16384,0.037512888511021934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,16384,0.041484445333480835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,256,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,256,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,12288,0.029719110992219713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,12288,0.031975110371907554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,8192,0.03606133328543769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,10240,0.025498666697078284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,10240,0.0277528895272149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,7168,0.03120533294147915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,8192,0.021909332937664453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,6144,0.027460444304678176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,7168,0.02161866592036353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,8192,0.03126222226354811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,5120,0.023920888702074688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,7168,0.026944888962639704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,6144,0.018227555685573153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,4096,0.019496889577971566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,6144,0.02417244513829549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,3584,0.01680088871055179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,5120,0.0158942225906584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,4096,0.014241778188281588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,3072,0.01459822224246131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,3584,0.01314133322901196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,4096,0.018914666440751817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,2560,0.01219288922018475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,51200,0.11271554893917507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,3584,0.015678221980730694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,51200,0.1023626658651564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,2048,0.010504888991514841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,3072,0.012340444657537671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,1536,0.00848711116446389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,3072,0.01629688839117686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,10240,0.0429004430770874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,1024,0.006904888898134232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,2048,0.009716444545321995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,2560,0.012998222476906247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,1536,0.008452444440788692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,768,0.005538666827811136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,1024,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,1536,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,2048,0.012336889074908363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,1024,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,512,0.003925333420435588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,256,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,768,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,128,0.0024764444679021835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,768,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,256,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,512,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,64,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,512,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,32,0.0022044444663657085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,128,0.0064222220745351576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,128,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,5120,0.021615111165576514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,64,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,65536,0.21678488784366182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,32,0.005978666659858491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,51200,0.1716817749871148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,16384,0.05810933642917209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,65536,0.1424257755279541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,12288,0.045178665055168994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,65536,0.1252222193611993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,2560,0.011426666544543372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,10240,0.037244445747799344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,16384,0.03945422172546387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,16384,0.03554133243030972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,51200,0.09855466418796116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,8192,0.03082488973935445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,12288,0.030461334519916113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,7168,0.027792887555228338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,12288,0.028157333532969158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,10240,0.024264888630972967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,8192,0.020851555797788832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,10240,0.026667555173238117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,8192,0.02641155487961239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,7168,0.019262222780121695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,5120,0.0211191111140781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,7168,0.022495110829671223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,4096,0.017596445149845548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,6144,0.01721155477894677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,6144,0.020336000455750358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,5120,0.015903110305468243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,3584,0.015127110812399121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,4096,0.0138568886452251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,5120,0.01791466606987847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,3072,0.013442666994200813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,4096,0.015896888242827523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,3584,0.01294844432009591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,2560,0.011384888655609555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,3584,0.014128888646761576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,3072,0.011571555501884885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,2048,0.009992889232105678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,2560,0.01035377797153261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,3072,0.013440888788965015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,1536,0.008084444536103142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,2560,0.011886222494973077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,1024,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,51200,0.10750933488210042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,1536,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,768,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,1536,0.009011555876996782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,1024,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,512,0.0038417776425679526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,6144,0.024417777856190998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,768,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,256,0.0028897778441508612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,768,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,512,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,128,0.002472888885272874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,256,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,512,0.006224888894293044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,64,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,256,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,32,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,128,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,128,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,65536,0.17347555690341523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,64,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,32,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,51200,0.12928356064690485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,2048,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,65536,0.09034666750166152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,16384,0.044376889864603676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,2048,0.011057777537239922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,65536,0.07282133234871759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,51200,0.06916444169150458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,12288,0.034832000732421875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,10240,0.02667377723587884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,1024,0.007382222347789341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,51200,0.05659200085533989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,12288,0.018333334061834548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,16384,0.027506666051016912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,8192,0.023649776975313824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,12288,0.021327111456129286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,7168,0.02068088948726654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,10240,0.016382222374280293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,6144,0.018219555417696636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,10240,0.01891377733813392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,7168,0.01353422221210268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,8192,0.01958755486541324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,7168,0.018231999542978074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,6144,0.012561777399645912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,4096,0.013065777719020844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,6144,0.016711999972661335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,5120,0.011614222493436603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,3584,0.011592888997660743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,4096,0.010416888528399997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,3072,0.010470222267839642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,3584,0.009600888523790572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,4096,0.013350221845838757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,3584,0.011685333318180509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,2560,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,3072,0.008936000367005667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,2048,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,3072,0.011220444407727985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,2560,0.008110221889283922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,1536,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,2560,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,2048,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,1024,0.0047280001971456735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,16384,0.02329866588115692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,2048,0.009204444785912832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,768,0.0038533334930737815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,1536,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,1024,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,8192,0.014718222949239941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,512,0.002944888960984018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,1024,0.006665777828958299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,5120,0.015486222174432544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,256,0.0024986666523747975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,768,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,768,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,128,0.002130666747689247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,512,0.006112888869312074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,512,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,256,0.005881777654091517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,64,0.0019093333846992916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,128,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,32,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,5120,0.014263111684057446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,128,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,65536,0.14856444464789495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,64,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,32,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,65536,0.08640711175070868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,65536,0.06598133511013456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,51200,0.06299822198020087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,51200,0.05395644572046068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,1536,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,16384,0.02551377813021342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,16384,0.021867555048730638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,10240,0.026779555612140234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,12288,0.01776977711253696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,12288,0.019916445016860962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,8192,0.023137778043746948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,10240,0.01572355628013611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,256,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,10240,0.01722577710946401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,7168,0.019362666540675692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,8192,0.01960000064637926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,8192,0.013987556099891663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,6144,0.01685599982738495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,7168,0.01297155519326528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,5120,0.014626666903495789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,7168,0.017490666773584154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,6144,0.012111110819710625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,4096,0.012865777644846173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,16384,0.040392001469930015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,6144,0.015498666299713982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,3584,0.011023999916182624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,5120,0.011038222246699862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,3072,0.009791110952695211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,4096,0.009582222335868413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,5120,0.01367377738157908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,2560,0.0086666668454806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,3584,0.00868622213602066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,3584,0.011065777805116443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,3072,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,2048,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,2560,0.007335111498832703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,3072,0.010818666881985135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,1536,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,2560,0.00962044464217292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,2048,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,1024,0.004553777890072929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,1536,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,2048,0.008527999950779809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,768,0.003604444364706675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,1536,0.007266666326257918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,1024,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,512,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,512,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,51200,0.11572178204854329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,128,0.002151111140847206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,256,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,256,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,64,0.0019093333846992916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,128,0.005639111002286275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,4096,0.012471111284361945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,32,0.0021235555824306277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,128,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,65536,0.12145155005984837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,64,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,32,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,51200,0.09563199679056804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,65536,0.07591378026538424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,512,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,65536,0.06360710991753472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,768,0.006069333189063602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,51200,0.05660533242755466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,12288,0.027129777603679236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,51200,0.0514977773030599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,16384,0.020949333906173706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,10240,0.023324444890022278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,256,0.002463111033042272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,8192,0.02001066671477424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,12288,0.017229333519935608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,12288,0.018186666899257235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,10240,0.015244444211324057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,7168,0.017647110753589206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,8192,0.013667555318938361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,6144,0.01573955516020457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,8192,0.018773333893881906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,7168,0.012764444781674279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,12288,0.031692445278167725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,5120,0.013783999615245394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,7168,0.018529777725537617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,4096,0.01075111081202825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,6144,0.011680888632933298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,6144,0.014418666561444601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,5120,0.010423111418883005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,3584,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,5120,0.013068444199032254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,4096,0.008957333034939235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,3072,0.009192888935407003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,4096,0.012426666915416718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,3584,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,16384,0.034212443563673235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,2560,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,3584,0.010407999985747868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,2048,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,3072,0.007802666889296637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,2560,0.007752888732486301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,3072,0.010339555641015371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,1536,0.005893333504597346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,16384,0.023769777682092454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,2048,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,1024,0.004148444367779626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,2048,0.007674666742483775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,2560,0.008746667040718926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,1536,0.006897777732875612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,10240,0.015986666083335876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,1024,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,512,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,768,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,256,0.0022479999396536085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,768,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,512,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,128,0.002191999927163124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,512,0.005631999837027655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,256,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,256,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,64,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,32,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,128,0.005630222045713001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,128,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,65536,0.10561422506968181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,64,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,32,0.005381333331267039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,51200,0.08395910925335354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,65536,0.06656889120737712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,65536,0.06119910875956217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,16384,0.0306968887646993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,51200,0.05129955543412102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,768,0.0031999999450312722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,51200,0.04944622185495165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,12288,0.024692444337738886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,1536,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,16384,0.020285333196322124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,16384,0.021278222401936848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,10240,0.02093066606256697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,8192,0.018240888913472492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,1024,0.006224888894293044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,12288,0.01679733395576477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,7168,0.01609155535697937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,10240,0.01482488877243466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,12288,0.016680000556839835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,6144,0.01421333352724711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,8192,0.013092444174819522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,10240,0.014831999937693277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,5120,0.012311110893885294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,8192,0.017855111095640395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,4096,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,7168,0.01566488875283135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,6144,0.011057777537239922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,3584,0.00943288869327969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,5120,0.009712888962692684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,6144,0.014133334159851074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,3072,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,4096,0.008572444319725037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,5120,0.012780444489585029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,2560,0.007528000407748752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,3584,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,4096,0.011510222322411008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,3072,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,2048,0.00655644428398874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,1536,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,3072,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,2560,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,1024,0.003699555579158995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,2560,0.007677333222495184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,1536,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,768,0.0032071111102898917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,2048,0.007626666790909237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,1024,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,1536,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,512,0.0028151110228565005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,1024,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,256,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,768,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,512,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,128,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,256,0.00564533347884814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,64,0.0018168888572189545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,256,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,128,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,32,0.001826666709449556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,128,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,65536,0.08441688617070515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,32,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,3584,0.010382222632567087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,65536,0.0599857767422994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,65536,0.05911733044518364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,2048,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,16384,0.025192888246642217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,51200,0.047600001096725464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,51200,0.04690222276581658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,12288,0.020047111643685233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,16384,0.01920355525281694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,16384,0.01960533360640208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,512,0.005387555393907759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,10240,0.01751377847459581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,12288,0.016190222567982145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,7168,0.012279111478063794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,8192,0.015792000624868605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,12288,0.015392889579137167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,10240,0.014426666829321118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,7168,0.013097777962684631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,10240,0.013795555465751223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,8192,0.012727110750145383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,6144,0.011349333657158745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,7168,0.01144177797767851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,8192,0.01687111126052009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,5120,0.010016889207892949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,7168,0.015293333265516492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,6144,0.010279999838935005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,4096,0.008904000123341879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,5120,0.00965244405799442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,6144,0.012719111310111152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,51200,0.06613510847091675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,3584,0.007962666451931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,3072,0.007076444725195567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,4096,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,5120,0.011872000164455838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,2560,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,4096,0.011038222246699862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,3584,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,3072,0.0074755557709270054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,2048,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,3072,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,1536,0.0039022221333450745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,2048,0.00675644435816341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,2560,0.007806222471925948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,1024,0.0031386667655573953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,2048,0.007085333267847697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,1536,0.0064106666379504735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,768,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,1536,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,1024,0.005982222242487802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,512,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,1024,0.005984000033802456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,768,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,64,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,768,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,128,0.0019786666250891155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,512,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,512,0.005283555636803309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,256,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,64,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,32,0.0021688888470331826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,256,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,128,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,65536,0.07724089092678495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,128,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,64,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,51200,0.0591191119617886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,32,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,3584,0.008524444368150499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,16384,0.02234399980968899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,65536,0.05811644262737698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,65536,0.0581351121266683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,12288,0.01790577835506863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,51200,0.046784000264273755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,10240,0.015483554866578845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,16384,0.01923111081123352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,16384,0.018186666899257235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,12288,0.015832889411184523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,8192,0.013239999612172445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,12288,0.014903111590279473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,7168,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,10240,0.013400000002649097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,8192,0.012170666621790992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,6144,0.010882666541470421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,256,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,7168,0.011131555669837527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,5120,0.00941955546538035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,7168,0.014009777042600842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,6144,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,4096,0.00869777798652649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,6144,0.011563555234008364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,5120,0.00906044410334693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,3584,0.007470221983061896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,5120,0.011682666838169098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,4096,0.008273777862389883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,3072,0.007037333316273159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,3584,0.007791111038790808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,2560,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,3584,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,2560,0.006179555422729916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,51200,0.04514133267932468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,3072,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,2048,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,2560,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,10240,0.014145778285132514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,1536,0.003866666720973121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,2560,0.0074177781740824384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,2048,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,2048,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,8192,0.01551555593808492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,1024,0.0031724443866146933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,768,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,1536,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,1536,0.005997333261701796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,1024,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,768,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,1024,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,256,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,512,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,512,0.005312888986534542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,128,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,4096,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,256,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,64,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,128,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,32,0.0018275556051068837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,3072,0.007440889047251807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,128,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,65536,0.06210222509172228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,64,0.005144888742102517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,32,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,51200,0.04927644464704725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,65536,0.055431998438305326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,65536,0.057265778382619224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,16384,0.0192284451590644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,51200,0.04593955477078756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,512,0.002502222235004107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,16384,0.01772177716096242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,768,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,10240,0.01516888870133294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,12288,0.015637333194414776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,12288,0.014475555883513557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,8192,0.013448889056841532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,10240,0.013773333695199756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,256,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,7168,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,8192,0.011696889168686338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,8192,0.01442488862408532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,6144,0.010987555815113915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,7168,0.010407110883129967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,7168,0.012869333227475485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,5120,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,6144,0.009472000102202097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,6144,0.011669333610269757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,4096,0.008267555799749162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,5120,0.008844444321261512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,5120,0.010006222460005019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,4096,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,4096,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,3072,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,3584,0.00795022232664956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,3584,0.007583111524581909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,2560,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,3072,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,3072,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,10240,0.013081777426931592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,2048,0.004559999952713649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,2560,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,1536,0.0038399998512532976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,2560,0.0070426662762959795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,2048,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,1024,0.0031902222997612427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,2048,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,1536,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,1536,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,768,0.0028719999310043124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,1024,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,512,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,1024,0.005656888915432825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,768,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,256,0.002128888956374592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,768,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,3584,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,128,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,512,0.0053484443989064955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,256,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,64,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,256,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,32,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,128,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,65536,0.05693688657548693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,64,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,32,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,51200,0.04377333323160807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,65536,0.056180444028642446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,65536,0.05317688981691996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,16384,0.019595555133289762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,51200,0.04530133472548591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,51200,0.041793776883019336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,16384,0.017287110288937885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,16384,0.018962666392326355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,512,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,16384,0.018696000178654987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,10240,0.014085332552591959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,12288,0.014816888504558139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,12288,0.014157333307796054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,128,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,8192,0.012768000364303589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,10240,0.012802666260136498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,10240,0.012739555703269111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,7168,0.011348444554540845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,8192,0.010956444674068026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,6144,0.010223111344708337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,7168,0.01014666673209932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,7168,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,5120,0.008739555875460306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,6144,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,6144,0.010395555860466426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,4096,0.007907555335097844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,5120,0.00871288859181934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,5120,0.008411555654472774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,3584,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,4096,0.007998222278224098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,12288,0.015829333000712924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,4096,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,3072,0.006091555373536215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,3584,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,2560,0.005311111195219888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,3072,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,3584,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,3072,0.007017778025733099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,2560,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,2560,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,1536,0.0037413334680928122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,2048,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,2048,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,8192,0.013077333569526672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,1024,0.0031386667655573953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,1536,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,768,0.0028071111689011254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,1536,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,1024,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,512,0.002528888897763358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,768,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,1024,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,256,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,768,0.005327111317051782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,128,0.0019173332386546666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,256,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,512,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,64,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,51200,0.04342577854792277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,128,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,32,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,128,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,65536,0.048003554344177246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,64,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,51200,0.03538577755292257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,32,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,65536,0.05535377727614509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,16384,0.016158221496476066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,65536,0.050620443291134305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,51200,0.04463111029730903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,51200,0.04033777780003018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,12288,0.01333333303531011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,16384,0.01830311119556427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,10240,0.011761777930789523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,16384,0.016784888174798753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,12288,0.014286221729384528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,8192,0.011044444309340583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,12288,0.013919111755159167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,10240,0.012148444851239523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,512,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,7168,0.009653333160612319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,10240,0.012400888734393649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,6144,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,8192,0.010732444624106089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,8192,0.011740444435013665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,7168,0.010042666561073726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,12288,0.016201777590645682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,7168,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,4096,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,5120,0.008760888543393876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,6144,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,3584,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,5120,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,2048,0.004405333350102107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,3072,0.005884444548024072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,4096,0.007684444387753804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,3584,0.007822222179836696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,2560,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,3584,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,3072,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,2048,0.004320000194840961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,3072,0.0069146665434042616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,2560,0.007332444190979004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,1536,0.0037831109431054858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,2560,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,2048,0.006903999795516332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,2048,0.006239111224810283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,1024,0.0031555555760860443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,1536,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,1536,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,768,0.0028204443968004654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,256,0.004961777892377642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,1024,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,6144,0.009650666680600908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,512,0.002549333290921317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,768,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,768,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,256,0.0021946666141351065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,512,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,128,0.0019964445382356644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,4096,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,512,0.005290666802061929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,256,0.005452444569932089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,64,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,32,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,256,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,128,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,65536,0.03453422254986233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,128,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,51200,0.027494221925735474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,64,0.004967111266321606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,32,0.004956444518433677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,16384,0.012661333713266583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,65536,0.05470488799942864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,65536,0.048651556173960366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,51200,0.04438933398988512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,51200,0.03882844580544366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,16384,0.017325333423084684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,10240,0.00998222248421775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,16384,0.0162124451663759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,12288,0.01310222182008955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,5120,0.008223999705579545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,12288,0.013456000222100152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,8192,0.008991999758614434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,7168,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,10240,0.012273777690198688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,10240,0.011714666253990598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,8192,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,6144,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,7168,0.010228444304731157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,8192,0.009917333722114563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,5120,0.007225777539942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,7168,0.009456000394291347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,6144,0.009493333597977957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,4096,0.006671999891599019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,6144,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,3584,0.005980444451173146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,4096,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,5120,0.008100444244013892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,3072,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,4096,0.007621333003044128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,3584,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,2560,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,3584,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,3072,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,12288,0.011242667006121742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,2048,0.004239999999602635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,3072,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,1536,0.003535110917356279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,2560,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,1024,0.0031964443624019623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,2048,0.006990222467316522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,2560,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,768,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,1536,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,1024,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,2048,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,512,0.002481777841846148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,1536,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,1024,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,256,0.002199999988079071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,768,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,1024,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,128,0.0018817778262827133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,512,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,768,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,256,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,32,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,128,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,512,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,256,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,65536,0.02777155571513706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,128,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,51200,0.021977777282396953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,5120,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,64,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,16384,0.010004444254769219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,65536,0.05420266588528951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,32,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,12288,0.008557333714432186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,51200,0.0436426666047838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,10240,0.007690666450394525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,16384,0.015856888559129503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,51200,0.03770133190684848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,12288,0.013249778085284762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,16384,0.0151582227812873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,12288,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,10240,0.01205599970287747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,7168,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,10240,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,8192,0.010703999963071613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,6144,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,8192,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,7168,0.010038221875826517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,5120,0.006315555423498154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,7168,0.009319999979601966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,6144,0.009319999979601966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,4096,0.006581333362393909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,6144,0.00870311094654931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,5120,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,64,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,3584,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,5120,0.00831911133395301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,3072,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,4096,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,3584,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,3072,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,65536,0.046931554873784385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,3072,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,2048,0.005807111246718301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,8192,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,2560,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,1536,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,2560,0.006320888797442119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,2048,0.006656000183688269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,2048,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,1024,0.0038906666967603895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,1536,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,768,0.0034328889515664843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,1536,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,1024,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,512,0.0028862222615215513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,1024,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,768,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,256,0.0022622222701708474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,768,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,512,0.005693333430422678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,512,0.004962666581074397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,4096,0.00812266684240765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,256,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,3584,0.00795377790927887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,64,0.0018862223045693503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,256,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,128,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,2560,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,65536,0.021724444296624925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,128,0.004957333207130432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,51200,0.01754755609565311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,64,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,32,0.004866666677925321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,16384,0.009434666898515489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,65536,0.053929779264662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,65536,0.04634933339224922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,51200,0.043453332450654775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,12288,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,51200,0.03714311122894287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,16384,0.015799111790127225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,10240,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,8192,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,16384,0.014494222071435718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,12288,0.01309866706530253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,7168,0.00656355544924736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,12288,0.012231999801264869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,8192,0.010447111394670276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,10240,0.011016888750924004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,6144,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,7168,0.010017777482668558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,5120,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,6144,0.009324444664849175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,128,0.0021671110557185277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,4096,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,6144,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,32,0.0018702221827374564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,5120,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,3584,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,5120,0.008072000410821702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,3072,0.005393777870469623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,4096,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,3584,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,2560,0.004867555366622077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,3584,0.007145778172545963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,3072,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,3072,0.0069857777820693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,2048,0.004222222086456087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,2560,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,1536,0.004469333423508538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,2560,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,2048,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,1024,0.003549333247873518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,2048,0.006079111248254776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,1536,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,10240,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,768,0.003128000017669466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,1536,0.005889777921968036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,512,0.0026577777332729767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,8192,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,768,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,768,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,7168,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,256,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,512,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,128,0.0018800000349680583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,512,0.005094222310516569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,256,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,4096,0.008179555336634317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,256,0.004852444347408083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,128,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,128,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,65536,0.015978667471143935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,64,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,51200,0.013105777402718862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,65536,0.05005688799752129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,65536,0.045548445648617215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,16384,0.007671111159854465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,51200,0.038443555434544883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,51200,0.036673777633243136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,12288,0.006717333363162146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,16384,0.015809777710172866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,1024,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,16384,0.014475555883513557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,10240,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,1024,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,8192,0.005976888868543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,12288,0.01308444473478529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,12288,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,7168,0.005401777724424998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,8192,0.010560000108347999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,8192,0.009734222458468543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,6144,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,768,32,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,7168,0.009999111294746399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,5120,0.0064222220745351576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,7168,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,6144,0.009309333231714036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,4096,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,5120,0.008661333057615492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,32,0.004616888860861461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,3584,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,4096,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,4096,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,3072,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,3584,0.0075866662793689305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,3584,0.0069004446268081665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,2560,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,3072,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,3072,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,2048,0.005024888863166173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,2560,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,2560,0.006464889066086874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,10240,0.01109866632355584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,1536,0.003547555456558863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,1024,0.0028773333049482773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,2048,0.00664533343580034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,1536,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,2048,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,768,0.0028071111689011254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,1024,0.005969777703285217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,512,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,1536,0.005882666756709416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,1024,0.0052986666560173035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,768,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,6144,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,768,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,5120,0.007666666474607255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,128,0.001863111017478837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,512,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,64,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,256,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,128,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,32,0.0018275556051068837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,65536,0.010503110786279043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,128,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,64,0.00462666650613149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,51200,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,32,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,65536,0.04618933465745714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,10240,0.011776888536082374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,16384,0.005281777845488654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,65536,0.04253866606288486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,12288,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,51200,0.03699466586112976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,16384,0.015815110670195687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,51200,0.034302221404181585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,10240,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,12288,0.013220444321632385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,16384,0.014497778481907315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,8192,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,10240,0.011702222128709158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,7168,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,512,256,0.0022008888837363985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,10240,0.010860444770918952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,8192,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,6144,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,512,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,8192,0.01001422190003925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,5120,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,7168,0.009884444375832876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,256,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,6144,0.00925422211488088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,4096,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,6144,0.008664888640244802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,5120,0.008631111019187504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,3584,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,4096,0.00816888858874639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,5120,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,3072,0.0046311111913787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,4096,0.007055111229419708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,3584,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,2560,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,3072,0.007415999968846639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,2048,0.0046675557063685525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,3072,0.006707555717892117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,2560,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,1536,0.003559111307064692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,2048,0.006904888898134232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,2048,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,1024,0.002871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,12288,0.012070222033394707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,1536,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,768,0.002566222308410539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,1024,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,1024,0.005684444473849402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,512,0.002516444358560774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,768,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,768,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,7168,0.008810666700204214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,256,0.0021199999998013177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,512,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,128,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,256,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,512,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,64,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,256,0.004629333400064045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,256,32,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,128,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,3584,0.007142222589916653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,65536,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,128,0.004861333303981357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,2560,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,64,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,65536,0.046013332075542875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,32,0.004560889055331548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,16384,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,65536,0.04178577661514282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,51200,0.03731911049948798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,12288,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,51200,0.03374222252104018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,16384,0.015824000040690105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,10240,0.005492444253630108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,16384,0.014158222410413953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,12288,0.013091555900043912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,1536,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,8192,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,7168,0.005097777893145879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,10240,0.011810666984981961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,12288,0.012263110942310758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,6144,0.005032888717121548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,8192,0.01071466671095954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,10240,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,5120,0.004596444467703502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,7168,0.010031110710567897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,8192,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,4096,0.004557333472702238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,6144,0.009335999687512716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,5120,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,3584,0.004638222356637319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,4096,0.00796977761718962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,5120,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,3072,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,4096,0.00737955586777793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,2560,0.004544888933499654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,3584,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,51200,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,3072,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,2048,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,3584,0.006688000013430913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,3072,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,1536,0.003547555456558863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,2560,0.00702133360836241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,1024,0.00281866660548581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,2048,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,2560,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,768,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,2048,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,1024,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,1536,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,512,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,768,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,1024,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,256,0.0021280000607172647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,768,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,512,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,128,0.0017964444640609955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,512,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,256,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,64,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,128,0.005506666584147348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,7168,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,128,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,6144,0.00850933293501536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,65536,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,51200,0.006796444455782573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,64,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,16384,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,32,0.00452888881166776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,12288,0.004941333499219683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,65536,0.042119999726613365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,10240,0.005457777943876054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,51200,0.03393866618474325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,8192,0.004960000101062986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,16384,0.014298665854665967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,7168,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,12288,0.011759111450778114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,1536,0.006325333482689328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,6144,0.004274666723277834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,5120,0.004536889079544279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,8192,0.009711110757456886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,4096,0.0041955556306574075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,7168,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,6144,0.008389333056079017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,3584,0.004572444491916233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,5120,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,4096,0.007360888852013483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,256,0.004577777865860197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,2560,0.004609777695602841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,3584,0.006694222076071634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,3072,0.006516444600290722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,1536,0.0032764443506797156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,2560,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,2048,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,1536,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,768,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,1024,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,512,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,10240,0.011076444553004371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,256,0.0019617778145604664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,768,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,128,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,512,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,64,0.001779555653532346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,3072,0.0042204442951414315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,32,0.0017991111510329777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,256,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,65536,0.008296888735559251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,128,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,128,32,0.001785777716173066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,51200,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,2048,0.00423911131090588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,16384,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,64,0.004573333180612988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,12288,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,10240,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,32,0.004857777721352047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,8192,0.004607999904288186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,65536,0.04213688770929972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,51200,0.03387022349569533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,64,1024,0.0028124445428450904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,16384,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,6144,0.00453244439429707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,5120,0.004584889031118817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,12288,0.011775111158688864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,4096,0.00388711111413108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,10240,0.011010666688283285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,8192,0.009720000127951304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,3584,0.004247111164861255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,3072,0.004222222086456087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,7168,0.009368889033794403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,5120,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,6144,0.008392888638708327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,2048,0.0041884444653987885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,3584,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,3072,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,1024,0.002875555513633622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,2560,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,768,0.0025031111306614345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,2048,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,512,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,1536,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,256,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,1024,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,128,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,768,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,64,0.0018071112119489247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,512,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,256,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,128,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,64,0.004952888935804367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,32,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,2560,0.004575110971927643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,4096,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,1536,0.0035111109415690103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,51200,0.48767466015285915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,16384,0.320395549138387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,16384,0.16287377145555285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,51200,0.9687235090467664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,32,0.001769777801301744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,13,32,7168,0.004597333156400257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,12288,0.2261271211836073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,16384,0.13675910896725124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,12288,0.12149155139923096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,10240,0.18803555435604521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,51200,0.43569244278801816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,12288,0.10395110978020562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,10240,0.12469689051310222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,10240,0.08687377638287014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,8192,0.15350843800438776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,8192,0.11324089103274876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,12288,0.1321395503150092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,16384,0.17509688271416557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,8192,0.07177866829766168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,7168,0.13268621762593588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,6144,0.11429954899681939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,7168,0.09921422269609238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,10240,0.11278400156233047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,7168,0.06350933180914985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,5120,0.09620355235205756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,8192,0.09592177470525105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,6144,0.055099556843439736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,4096,0.08105599880218506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,7168,0.08379555410808986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,5120,0.07649155457814534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,5120,0.04690400097105238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,3584,0.06820088624954224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,6144,0.07340621948242188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,4096,0.06038310792711046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,5120,0.06354666418499418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,3072,0.05910044246249729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,3584,0.05272355675697327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,4096,0.053822222683164805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,3072,0.030955556366178725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,3072,0.04935200015703837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,2560,0.049639999866485596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,6144,0.08690310849083795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,3072,0.04343466626273262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,2048,0.0409457782904307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,2560,0.039176887936062284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,2560,0.02772888872358534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,2560,0.03856177793608771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,2048,0.03675377700063918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,4096,0.03915289044380188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,1536,0.0314408904976315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,2048,0.0278479986720615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,2048,0.03422222203678555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,1024,0.022817777262793645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,3584,0.03503377901183234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,1536,0.0188773340649075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,1536,0.026701332794295415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,768,0.017881777551439073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,1024,0.015121777852376303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,1024,0.02067466742462582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,3584,0.0488302210966746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,512,0.012587555580668978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,768,0.013344888885815939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,51200,0.5048382017347548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,1024,0.022987556126382615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,256,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,512,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,512,0.01276000009642707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,128,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,256,0.010507555471526252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,512,0.016744888491100736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,256,0.010103111465771994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,64,0.004217777815130022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,128,0.008191111187140146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,128,0.010518222219414182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,32,0.004192000048028098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,64,0.010264000131024255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,32,0.009974222216341231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,1536,0.029149333635965984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,768,0.016102222932709586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,768,0.02034133341577318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,65536,0.515499538845486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,65536,1.0912870830959744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,256,0.014448889427714877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,51200,0.3904213375515408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,128,0.009944000177913243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,51200,0.8444533348083496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,16384,0.1306942171520657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,65536,0.4365244441562229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,12288,0.20955822202894422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,16384,0.11078666978412205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,12288,0.09804799821641709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,51200,0.3525991174909804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,10240,0.172343995836046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,16384,0.2759724458058675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,12288,0.08533866537941827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,8192,0.13962311214870876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,10240,0.10303733083936904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,10240,0.0694355567296346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,8192,0.08991555372873943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,12288,0.1062506635983785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,8192,0.05652444230185615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,7168,0.12291644679175483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,10240,0.0928071141242981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,7168,0.07747911082373725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,8192,0.07800444629457262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,7168,0.05209155546294319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,6144,0.10506844520568848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,6144,0.07173333565394084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,7168,0.06872622172037761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,16384,0.14166933960384792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,5120,0.08907910850312974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,6144,0.04420622189839681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,6144,0.060886224110921226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,5120,0.06247911188337538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,4096,0.07201511330074735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,5120,0.03935822182231479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,3584,0.06274222003089057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,5120,0.05340977840953403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,4096,0.031932443380355835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,4096,0.057886222998301186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,3072,0.054832001527150474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,3584,0.043008890416887074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,3584,0.02884622083769904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,4096,0.04521066612667508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,2560,0.04613066713015238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,3072,0.025560888979170058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,3584,0.041476445065604314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,3072,0.04066133168008592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,2048,0.0373253325621287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,2560,0.022878222995334204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,3072,0.03735555542839898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,2560,0.03202755583657159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,1536,0.02910311023394267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,2048,0.02021511064635383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,2560,0.03419999943839179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,65536,0.5175075531005859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,2048,0.030518223841985066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,1024,0.02054755555258857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,1536,0.016180444094869826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,1536,0.02234933276971181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,2048,0.030248890320460003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,768,0.015132443772421943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,1024,0.013265777793195514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,1024,0.017142222987280954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,512,0.010638222098350525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,768,0.01179377817445331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,768,0.014075555735164218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,1024,0.0199991116921107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,256,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,51200,0.398926231596205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,512,0.011212444139851464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,512,0.011424889167149862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,768,0.01822933389080895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,128,0.003506666670242945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,256,0.008348444269763099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,64,0.0027786667148272195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,512,0.015810666812790763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,128,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,128,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,128,0.009347555538018545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,64,0.008961777720186446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,32,0.008723555339707268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,65536,0.3698835637834337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,256,0.0122079998254776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,1536,0.02507200009293026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,65536,0.1724675496419271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,51200,0.2701653374565972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,51200,0.13124799728393555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,65536,0.1605866617626614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,16384,0.08932444122102524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,256,0.009306666751702627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,51200,0.12535377343495688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,32,0.0028693332440323303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,16384,0.04834844337569343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,12288,0.06959022416008843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,16384,0.04215555720859104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,10240,0.05308977762858073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,12288,0.03640355666478475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,16384,0.05576799975501167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,8192,0.04384711053636339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,10240,0.042259554068247475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,10240,0.03646488984425863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,12288,0.04522044459978739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,7168,0.03716355562210083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,8192,0.034381333324644305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,8192,0.02422222163942125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,51200,0.13693422741360134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,7168,0.03158311049143473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,8192,0.035419556829664446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,7168,0.022039999564488728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,6144,0.03289066751797994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,65536,0.17724622620476616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,5120,0.0281368891398112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,6144,0.019766221443812054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,7168,0.03137955400678847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,6144,0.02896711230278015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,4096,0.02368533280160692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,5120,0.02551288902759552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,5120,0.01754399968518151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,3584,0.020374221934212577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,4096,0.01587733295228746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,4096,0.02130488885773553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,3584,0.014148443937301636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,4096,0.022429333792792425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,3584,0.01828888886504703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,3072,0.017704889178276062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,10240,0.038944890101750694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,3584,0.02107644412252638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,2560,0.014875556031862894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,3072,0.01740800009833442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,3072,0.013076444466908773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,3072,0.019314666589101154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,2048,0.01203644441233741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,2560,0.014484445254007975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,2560,0.017960889471901786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,2048,0.013618666264745923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,6144,0.029669334491093952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,1536,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,12288,0.03381333417362637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,5120,0.02530755599339803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,2048,0.016813332835833233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,1024,0.007658667034573025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,1536,0.01075111081202825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,1536,0.011711999773979187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,768,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,1024,0.007979555262459649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,1536,0.015480889214409722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,1024,0.009712888962692684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,512,0.004509333521127701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,1024,0.013570666313171387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,768,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,768,0.007237333390447829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,256,0.002536888958679305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,512,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,2560,0.011996444728639392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,512,0.006719111154476802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,128,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,768,0.01237511138121287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,256,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,512,0.01074044406414032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,64,0.002130666747689247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,128,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,128,0.006663110935025745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,32,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,2048,0.010989333192507425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,128,0.008731555607583787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,64,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,65536,0.25240710046556264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,32,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,65536,0.14806311660342747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,51200,0.19809155993991426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,65536,0.13578222857581243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,51200,0.1125040054321289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,16384,0.06530933247672187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,256,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,16384,0.04146755403942532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,12288,0.05050577720006307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,16384,0.036928001377317644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,256,0.01107911103301578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,16384,0.04464800159136454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,10240,0.042972445487976074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,12288,0.03172711200184292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,12288,0.029190222422281902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,8192,0.035628444618648954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,12288,0.03649244374699063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,51200,0.10690755314297146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,10240,0.03075288732846578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,65536,0.13882044951121011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,7168,0.031255112753974065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,8192,0.031204445494545832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,8192,0.021896888812383015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,10240,0.03258666727277968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,6144,0.02740977704524994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,7168,0.028083556228213843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,7168,0.01958400011062622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,8192,0.02976177798377143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,5120,0.02381955583890279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,6144,0.024949333733982507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,6144,0.018000000052981906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,7168,0.02641777859793769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,4096,0.019341333044899833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,51200,0.10687555869420369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,5120,0.01590755581855774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,5120,0.02232355541653103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,3584,0.01649599936273363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,6144,0.024280000064108107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,4096,0.014528888795110913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,4096,0.020167110694779288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,3072,0.014759999182489185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,5120,0.021959111094474792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,3584,0.01310755560795466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,3584,0.016523554921150208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,2560,0.012422222230169507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,3072,0.012059555285506778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,10240,0.037081778049468994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,3584,0.019310222731696237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,3072,0.015980444020695157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,2048,0.01052177780204349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,2560,0.01348533315791024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,3072,0.01754133403301239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,1536,0.008534222013420528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,2048,0.009710222482681274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,2560,0.01662400033738878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,2048,0.01255644443962309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,1024,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,1536,0.008657777474986183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,1536,0.01036088913679123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,2048,0.01574222246805827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,768,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,1024,0.007710222568776872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,1536,0.014408888088332282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,1024,0.00829777783817715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,512,0.003867555409669876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,768,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,768,0.007154666715198093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,1024,0.011706666813956367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,256,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,512,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,512,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,768,0.011187555889288584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,128,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,256,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,2560,0.01109866632355584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,512,0.010468444062603844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,64,0.0021946666141351065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,128,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,128,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,256,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,32,0.0022026666750510535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,64,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,128,0.008279999925030602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,65536,0.21642933951483834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,32,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,65536,0.14182666937510172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,51200,0.1712319983376397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,65536,0.12463911374409993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,51200,0.10767377747429742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,16384,0.057927111784617104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,4096,0.019663110375404358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,51200,0.10321422417958577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,256,0.005675555517276128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,16384,0.039517333110173546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,12288,0.045280890332327954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,16384,0.03952711158328586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,10240,0.03704444567362467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,12288,0.028421332438786823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,16384,0.04099377658632066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,12288,0.030720889568328857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,8192,0.030532442861133154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,51200,0.09406222237481011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,10240,0.027526222997241553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,10240,0.031147556172476873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,7168,0.0277511113219791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,8192,0.026529777381155226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,8192,0.020928000410397846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,10240,0.030014223522610132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,6144,0.024393777052561443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,7168,0.018985778093338013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,8192,0.02657866643534766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,7168,0.02318488889270359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,5120,0.020967110991477966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,6144,0.01721155477894677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,7168,0.02513422237502204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,4096,0.017647999856207106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,5120,0.015449777245521545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,5120,0.018583999739752877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,6144,0.022872888379626807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,3584,0.01514844430817498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,4096,0.016050666570663452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,4096,0.013898667362001208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,65536,0.12172444661458333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,3072,0.013568888935777875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,12288,0.03263555632697211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,3584,0.012882666455374824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,3584,0.014269333746698169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,4096,0.01848444508181678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,2560,0.011406222151385413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,3072,0.011735999749766456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,3584,0.018176888426144917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,2048,0.009666666388511658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,3072,0.013752889302041797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,2560,0.011941333611806234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,2560,0.010686222049925061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,1536,0.008224888808197444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,3072,0.016526222229003906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,2048,0.009091555244392818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,2560,0.015820443630218506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,2048,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,6144,0.020626667473051283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,1536,0.00813244448767768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,1536,0.009335999687512716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,2048,0.015139554937680563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,768,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,1024,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,1536,0.013975999421543546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,512,0.0041653331783082755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,5120,0.02026844521363576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,768,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,768,0.006806222101052602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,1024,0.012039110892348819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,256,0.002847111059559716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,512,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,512,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,768,0.011066666907734342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,128,0.0025395556456512874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,256,0.005658666706747479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,256,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,512,0.011002666420406766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,64,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,256,0.010220444036854638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,128,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,32,0.0021928888228204516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,128,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,64,0.005886222339338726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,1024,0.00639111093348927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,65536,0.17299021614922416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,128,0.008976000050703684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,32,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,51200,0.12883644633822972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,65536,0.0907671120431688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,1024,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,16384,0.044237332211600415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,65536,0.07014666663275824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,51200,0.06868977679146661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,12288,0.03465066684616937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,51200,0.05871555540296766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,16384,0.02254044512907664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,16384,0.02733777794573042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,10240,0.02682400080892775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,12288,0.018372444642914664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,12288,0.021367111139827307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,16384,0.037816888756222196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,51200,0.08955999877717759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,8192,0.02414488957987891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,65536,0.11100622018178304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,10240,0.023753777146339417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,10240,0.01646933290693495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,7168,0.020801777640978496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,12288,0.03221688999070062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,8192,0.01514400045077006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,6144,0.01812355551454756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,8192,0.020271999968422782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,7168,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,10240,0.02864177690611945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,5120,0.015672889020707872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,7168,0.018529777725537617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,8192,0.025105777713987563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,7168,0.023392000132136877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,4096,0.01330933305952284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,5120,0.01480888823668162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,5120,0.011671110987663269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,6144,0.021306667062971327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,3584,0.011379555695586734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,5120,0.020231111182106864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,4096,0.010401777923107147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,4096,0.013365333278973898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,3072,0.010346666806273991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,3584,0.009546666509575313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,3584,0.011770666473441653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,4096,0.018235555953449674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,2560,0.008796444369686974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,3072,0.011406222151385413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,3072,0.008964444200197855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,3584,0.017834666702482436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,2048,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,2560,0.008352888955010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,2560,0.01033688916100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,1536,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,3072,0.016864889197879367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,2048,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,2560,0.01566133399804433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,1024,0.0047208890318870544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,6144,0.012768000364303589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,1536,0.007053333024183909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,2048,0.014781332678265043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,1536,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,768,0.0038444445365005066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,1024,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,1024,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,512,0.0028391112056043413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,1536,0.013241777817408243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,768,0.00608800020482805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,1024,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,256,0.0026782221264309357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,512,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,512,0.0057297779454125305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,256,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,256,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,512,0.010442666709423065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,256,0.0099973330895106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,128,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,2048,0.009300444689061906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,64,0.002003555496533712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,32,0.0021608887861172357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,128,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,128,0.008623110751310984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,6144,0.01648622254530589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,65536,0.14876533879174128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,64,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,32,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,51200,0.1154337724049886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,768,0.006319111006127463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,65536,0.08304266797171699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,65536,0.06580888562732272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,16384,0.0402728882100847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,51200,0.06279733445909289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,51200,0.05367111166318258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,128,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,12288,0.031071109904183283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,16384,0.0255039996571011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,16384,0.02217866645918952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,65536,0.11067110962337917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,10240,0.026622222529517278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,16384,0.03804355528619554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,51200,0.08928266498777603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,12288,0.017838221457269456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,12288,0.020230222079488967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,8192,0.023053333163261414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,10240,0.015795555379655626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,10240,0.022613333331214056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,12288,0.030974222554100886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,7168,0.01958666741847992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,8192,0.01964977714750502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,8192,0.014089778065681458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,10240,0.02812444501452976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,6144,0.01699555582470364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,8192,0.02499466637770335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,7168,0.0129466669427024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,7168,0.018183999591403537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,5120,0.014659555421935188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,768,0.010678222609890832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,6144,0.011968000067604912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,6144,0.015976889265908133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,4096,0.012639111114872826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,5120,0.010899555351999072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,5120,0.014142221874660917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,6144,0.021661332911915247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,3584,0.011003555523024665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,4096,0.009677333136399588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,4096,0.012424888710180918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,5120,0.019401777121755812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,3072,0.00978311151266098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,3584,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,4096,0.018089777893490262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,2560,0.00830844458606508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,3072,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,3584,0.017031111650996737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,3072,0.016894222961531747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,2560,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,2048,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,2560,0.009683555199040307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,2048,0.00699733363257514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,2560,0.015447111593352424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,1536,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,2048,0.008984888593355814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,7168,0.022537777821222942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,1024,0.004514666481150521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,2048,0.015154666370815702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,1536,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,1536,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,768,0.0038844446341196695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,1024,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,1024,0.006367110957702001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,512,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,3584,0.011413333316644033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,768,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,768,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,1024,0.011894221935007306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,256,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,3072,0.01089599976936976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,512,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,128,0.0021484444538752237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,512,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,768,0.011003555523024665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,256,0.00573333352804184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,256,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,64,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,32,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,512,0.01108888867828581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,128,0.005658666706747479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,128,0.004867555366622077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,256,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,65536,0.1211191150877211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,64,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,128,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,32,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,51200,0.09570044279098511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,16384,0.0340266658200158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,65536,0.06355733341640897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,51200,0.05674844317966037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,51200,0.05173155665397644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,1536,0.012082666986518435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,16384,0.023149333066410486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,16384,0.02070488863521152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,65536,0.10965244637595283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,51200,0.08904977639516194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,16384,0.037335998482174344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,10240,0.023253333237436082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,12288,0.017254221770498488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,8192,0.01997955474588606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,10240,0.015217777755525378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,10240,0.0206977774699529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,12288,0.03148444493611654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,8192,0.018391999933454726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,7168,0.01762488815519545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,10240,0.027015111512607996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,8192,0.013750221994188098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,8192,0.024329778220918443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,65536,0.07575377490785387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,6144,0.01573333309756385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,7168,0.016508445143699646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,6144,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,12288,0.026986666851573523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,5120,0.013811555173661975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,6144,0.015410666664441427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,6144,0.02083911167250739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,12288,0.018212444252438016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,5120,0.010641777680979835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,4096,0.010681777364677854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,4096,0.009005332986513773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,5120,0.019524445136388142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,3584,0.009264888862768808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,4096,0.012127999630239276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,3584,0.00849422232972251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,4096,0.017828444639841717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,3584,0.010980444649855295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,3072,0.009269333548016017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,7168,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,3584,0.01682488951418135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,3072,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,7168,0.022317333353890315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,3072,0.016520889268981088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,2560,0.009386666946940953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,2048,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,2560,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,5120,0.014136888914638095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,1536,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,2048,0.007903111477692923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,2560,0.015365334020720588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,2048,0.01370044466522005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,1536,0.00666133314371109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,1536,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,1536,0.01202755586968528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,1024,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,768,0.0032391111469931076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,2560,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,1024,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,1024,0.011402666568756104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,512,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,768,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,768,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,256,0.002444444431198968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,512,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,768,0.011056000159846412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,2048,0.007098666495747036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,256,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,512,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,128,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,1024,0.0042160000238153665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,128,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,64,0.0018986666368113623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,32,0.0018995555324686898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,128,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,65536,0.105712890625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,64,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,32,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,3072,0.007928888830873702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,51200,0.08401866753896077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,65536,0.06124177906248304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,512,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,51200,0.051391111479865186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,16384,0.030538668235143025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,65536,0.1099475555949741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,51200,0.049752887752321034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,256,0.010013333625263637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,16384,0.02125066684352027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,51200,0.08900533119837443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,12288,0.024538666009902954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,16384,0.02015911042690277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,10240,0.020862221717834473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,12288,0.016537777251667447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,128,0.008955555657545725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,16384,0.03714399867587619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,8192,0.018227555685573153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,12288,0.016570667425791424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,10240,0.02052622205681271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,10240,0.014653333359294467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,7168,0.015855999456511605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,12288,0.030731555488374498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,8192,0.013229333692126803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,65536,0.06769066386752658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,6144,0.01423733267519209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,10240,0.026756443911128577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,7168,0.012429333395428128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,8192,0.023746665981080797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,5120,0.012121777567598553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,6144,0.011388444238238864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,256,0.0057848890622456866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,6144,0.014344889256689282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,4096,0.01089244418674045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,6144,0.020616888999938965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,5120,0.009639999932712978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,5120,0.013087111214796701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,4096,0.008623110751310984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,4096,0.011741333537631564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,5120,0.01889866590499878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,4096,0.017369776964187622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,3584,0.010327999790509542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,3072,0.008616000413894653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,3584,0.008021333151393467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,8192,0.017802667286660936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,3584,0.016555555992656283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,2560,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,7168,0.015564444992277356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,3072,0.009888889061080085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,7168,0.022680888573328655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,2560,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,3072,0.01516711049609714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,2560,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,2560,0.015121777852376303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,2048,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,1536,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,2048,0.007679999702506595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,3584,0.009741333623727163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,1024,0.0038577777643998465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,2048,0.01290933373901579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,1536,0.006663110935025745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,1536,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,768,0.0032684444967243406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,1024,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,1024,0.005969777703285217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,1536,0.0123848890264829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,512,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,768,0.00600622221827507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,1024,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,768,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,3072,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,256,0.0023173333870040043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,512,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,512,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,768,0.0107004443804423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,128,0.0021377777059872947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,2048,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,256,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,64,0.0018568889548381169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,512,0.010715555813577441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,32,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,128,0.005646222167544895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,128,0.004954666727119022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,256,0.009703111317422655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,65536,0.08466044399473403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,64,0.005390222287840313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,32,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,51200,0.06619199779298571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,65536,0.060772445466783315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,65536,0.059055997265709766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,16384,0.02517066730393304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,51200,0.046895109944873385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,51200,0.047650665044784546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,12288,0.02020088831583659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,65536,0.10998577541775173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,16384,0.019288000133302476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,16384,0.01958933307064904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,51200,0.0885431104236179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,10240,0.017647110753589206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,12288,0.01610844499535031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,12288,0.015321777926550971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,8192,0.015389333168665567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,16384,0.03637511200375027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,256,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,10240,0.014335999886194864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,7168,0.012827555338541666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,12288,0.029520889123280842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,8192,0.012488889197508493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,10240,0.026760000321600173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,8192,0.016809778081046212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,6144,0.011218667030334473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,128,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,7168,0.011542222566074796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,7168,0.014504889647165934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,5120,0.010058666268984476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,8192,0.023638221952650283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,6144,0.01038577738735411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,4096,0.009298666483826106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,5120,0.009083555804358589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,5120,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,4096,0.010984000232484607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,5120,0.018751111295488145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,3584,0.00793866647614373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,4096,0.008340444829728868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,3584,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,4096,0.0173662222094006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,3072,0.006989333364698622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,3584,0.009725333087974125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,10240,0.018245332770877414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,2560,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,3584,0.016466667254765827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,3072,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,3072,0.008384889198674096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,2048,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,2560,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,3072,0.015821332732836407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,2560,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,2048,0.007133333219422235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,6144,0.014133334159851074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,1536,0.004264889078007804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,2560,0.013833777772055732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,7168,0.02198666665289137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,1024,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,6144,0.020590222544140287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,1536,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,2048,0.01275288893116845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,1536,0.00666133314371109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,1024,0.006247111078765657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,1024,0.006222222414281633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,1536,0.012078222301271228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,512,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,768,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,1024,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,512,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,768,0.010678222609890832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,256,0.002190222135848469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,512,0.01034400032626258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,256,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,2048,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,256,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,128,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,256,0.010051555931568146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,64,0.0018702221827374564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,128,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,32,0.0021280000607172647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,768,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,128,0.004962666581074397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,65536,0.07685244745678373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,128,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,64,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,32,0.004944889081848992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,51200,0.05881600247489082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,768,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,65536,0.05780533287260267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,16384,0.022320000661744013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,65536,0.058415108256869845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,512,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,51200,0.04667644368277656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,51200,0.045032888650894165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,12288,0.018033777674039204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,16384,0.019336000084877014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,16384,0.018350222044520907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,10240,0.01551555593808492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,12288,0.01589866644806332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,51200,0.08869600296020508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,12288,0.01480977733929952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,8192,0.013316444224781461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,16384,0.036335110664367676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,10240,0.014168888330459595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,10240,0.017877333694034152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,7168,0.012078222301271228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,12288,0.029442667961120605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,65536,0.10971199803882176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,8192,0.012436444560686747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,8192,0.01514400045077006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,7168,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,7168,0.014127110441525778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,5120,0.009617778162161509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,8192,0.023984000086784363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,6144,0.010326222413116032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,6144,0.012840888566441007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,7168,0.021689777572949726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,4096,0.008621333373917473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,5120,0.008984000318580205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,6144,0.020626667473051283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,5120,0.011761777930789523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,3584,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,4096,0.008020444048775567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,4096,0.00979911122057173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,5120,0.019211555520693462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,3584,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,3584,0.007788444558779399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,4096,0.016904888881577384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,2560,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,3072,0.007332444190979004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,3584,0.016283555163277518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,2560,0.0069919998447100324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,6144,0.011009777585665384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,2048,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,3072,0.01477777792347802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,2560,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,1536,0.00387999994887246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,2048,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,2560,0.013285333083735572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,2048,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,1024,0.0031715554909573663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,1536,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,1536,0.006303999986913469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,2048,0.01277511070171992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,768,0.002879111096262932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,1024,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,1024,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,1536,0.012402666939629449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,512,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,768,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,1024,0.011015111373530494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,3072,0.008645333349704742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,512,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,512,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,256,0.002161777681774563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,768,0.010683555569913654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,10240,0.026208889153268602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,256,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,128,0.0021582223061058256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,512,0.010053333308961656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,256,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,64,0.0018542222678661346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,128,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,256,0.010488889283604093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,128,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,64,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,65536,0.06198222107357449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,128,0.008330666356616551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,32,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,3072,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,51200,0.04895377821392483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,65536,0.05720800161361694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,65536,0.05500444438722399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,768,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,16384,0.01917777789963616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,51200,0.04610133171081543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,65536,0.10972711112764146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,16384,0.01771377854877048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,16384,0.019063111808564927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,51200,0.08852444754706489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,16384,0.035649776458740234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,10240,0.015040889382362366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,12288,0.015493333339691162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,32,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,12288,0.02951288885540432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,10240,0.013857777747843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,8192,0.013215111361609565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,8192,0.012100444071822695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,10240,0.02674755619631873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,7168,0.01202666676706738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,8192,0.013621333572599622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,7168,0.01035911093155543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,8192,0.023623999622133043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,7168,0.01218311074707243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,51200,0.043248888519075185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,6144,0.009759110709031424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,12288,0.01573955516020457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,5120,0.008948444492287105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,6144,0.012344888514942594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,6144,0.020693333612547982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,5120,0.008956444760163626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,4096,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,5120,0.010689777632554373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,5120,0.018797333041826885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,4096,0.008299555215570662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,3584,0.007171555525726742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,4096,0.008616888688670265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,4096,0.01754488878779941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,3584,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,3072,0.006325333482689328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,3584,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,6144,0.010973333484596677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,3072,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,2560,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,3072,0.00777599960565567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,7168,0.021689777572949726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,2048,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,2560,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,2560,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,12288,0.014868444866604276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,1536,0.0038817777401871155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,2048,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,2048,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,10240,0.016515556308958266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,1536,0.006679111056857639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,2048,0.012776888906955719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,1024,0.003170666595300039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,1536,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,768,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,1024,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,1024,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,1536,0.012383999923865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,512,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,768,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,768,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,1024,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,3584,0.015474667151769003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,256,0.002151111140847206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,512,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,768,0.010664888554149203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,512,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,128,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,3072,0.014857777290874057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,256,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,256,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,64,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,512,0.010370666782061258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,32,0.0018293333964215384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,128,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,256,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,65536,0.05584000216590034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,2560,0.013737777868906656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,128,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,51200,0.04349422123697069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,64,0.004960888789759742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,32,0.004952888935804367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,128,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,65536,0.05612888601091173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,16384,0.019651555352740817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,51200,0.04525244567129347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,51200,0.04173244370354546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,12288,0.01633777717749278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,16384,0.01739644507567088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,16384,0.018684445155991446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,51200,0.08823111322191025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,10240,0.013888888888888888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,16384,0.035529779063330755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,12288,0.014931556251313953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,12288,0.014112000664075216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,8192,0.01275199982855055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,10240,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,12288,0.02938844429122077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,10240,0.015156444576051501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,7168,0.011223110887739392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,8192,0.011413333316644033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,10240,0.026085333691702947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,8192,0.013010666602187686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,6144,0.010056888891590966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,7168,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,8192,0.023406222462654114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,65536,0.052982221047083534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,5120,0.008774222599135505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,6144,0.009466666314336989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,7168,0.021576889687114294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,65536,0.10958311292860244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,4096,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,6144,0.020237333244747586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,5120,0.008957333034939235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,3584,0.00701155596309238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,4096,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,5120,0.018242667118708294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,4096,0.0075937774446275495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,3072,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,3584,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,3584,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,4096,0.016175111134847004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,2560,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,3072,0.0073928890956772705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,3072,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,3584,0.014833778142929077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,2048,0.004523555437723796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,7168,0.010447999669445885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,2560,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,2560,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,3072,0.014489778214030795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,1536,0.0038204445607132385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,2048,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,2048,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,6144,0.010753778119881948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,1024,0.003152888889114062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,2560,0.013440888788965015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,1536,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,1536,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,768,0.002833777831660377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,2048,0.012432888978057437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,1024,0.005991999887757831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,512,0.0025191110455327565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,1536,0.012147555748621622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,1024,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,768,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,256,0.0021502222451898786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,768,0.005231110999981563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,1024,0.011061333119869232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,512,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,128,0.0018817778262827133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,512,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,768,0.010729778144094678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,256,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,64,0.0018026667336622875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,512,0.010672888822025724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,32,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,128,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,256,0.010451555252075195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,65536,0.049660444259643555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,64,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,128,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,32,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,51200,0.034918222162458636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,65536,0.05538577834765116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,65536,0.049978666835361056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,16384,0.01616533266173469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,5120,0.008390222158696916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,51200,0.04019555449485779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,51200,0.044623110029432506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,12288,0.01328888866636488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,65536,0.10964977741241455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,16384,0.01683022247420417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,16384,0.01810933318403032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,51200,0.08784266975190903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,10240,0.011701333026091257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,12288,0.013975999421543546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,12288,0.013765333427323235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,16384,0.03569422165552775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,256,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,8192,0.010688888529936472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,10240,0.014469333820872836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,10240,0.012107555237081317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,128,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,7168,0.009653333160612319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,12288,0.029348442951838177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,10240,0.0262444449795617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,8192,0.011066666907734342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,6144,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,8192,0.012395555774370829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,7168,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,5120,0.008019555773999956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,7168,0.011394666300879585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,8192,0.02348088887002733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,6144,0.009376889301670922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,4096,0.0075795559419526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,7168,0.02193333374129401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,5120,0.008683555656009251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,5120,0.008221333225568136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,6144,0.019567110472255282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,3584,0.006863111009200414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,4096,0.008034666379292807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,4096,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,5120,0.017846221725145977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,3072,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,3584,0.007695111135641734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,4096,0.016149333781666227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,3584,0.007093333535724216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,2560,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,3072,0.0069857777820693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,3072,0.007511999871995714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,2048,0.004223111189073986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,3584,0.015207111835479736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,2560,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,3072,0.013789333403110504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,2560,0.006581333362393909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,1536,0.0037173334923055437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,2048,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,2048,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,2560,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,1024,0.0031964443624019623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,1536,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,1536,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,2048,0.01276000009642707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,6144,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,1536,0.011762667033407422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,1024,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,512,0.002487999904486868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,768,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,1024,0.011406222151385413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,768,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,256,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,512,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,768,0.010773333410422007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,256,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,256,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,128,0.0019351111518012153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,512,0.010316444767846001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,128,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,256,0.010332444475756751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,32,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,128,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,768,0.002842666581273079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,64,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,65536,0.034797334008746676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,32,0.004872888740566042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,1024,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,51200,0.02755022214518653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,65536,0.0547599991162618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,16384,0.012551999754375882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,65536,0.04880977670351664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,51200,0.04414933257632785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,51200,0.039064887497160174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,12288,0.01167555567291048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,512,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,65536,0.10911199781629775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,16384,0.01720355616675483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,10240,0.009748444788985783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,16384,0.016196444630622864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,51200,0.08800711234410603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,12288,0.013282666603724161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,12288,0.013444444371594323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,8192,0.008995555341243744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,16384,0.03525866733656989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,10240,0.012095111111799875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,10240,0.012104888757069906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,7168,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,12288,0.029463112354278564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,128,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,8192,0.010727110836240979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,6144,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,8192,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,10240,0.02660711109638214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,5120,0.007038222418891058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,7168,0.009444444543785518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,8192,0.022487999664412603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,6144,0.009314667019579146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,6144,0.008816888762844933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,7168,0.020614221692085266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,4096,0.0069164443347189165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,5120,0.008789333204428354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,5120,0.008204444415039485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,3584,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,6144,0.01954844428433312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,4096,0.008310221963458592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,5120,0.017435555656750996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,4096,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,3072,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,3584,0.008013333711359236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,3584,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,4096,0.01575911045074463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,2560,0.004861333303981357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,3072,0.007667555577225155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,3584,0.015167999598715039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,3072,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,2048,0.004234666625658671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,2560,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,3072,0.014118222726715935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,1536,0.0035422220826148987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,2048,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,2048,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,7168,0.010002666877375709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,1024,0.0031786666562159858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,1536,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,1536,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,2048,0.012709333664841123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,768,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,1024,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,1024,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,1536,0.011746666497654386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,512,0.0025351111673646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,768,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,768,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,1024,0.011369778050316704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,256,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,512,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,512,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,768,0.011408000356621213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,128,0.0018782222436534034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,256,0.005382222019963794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,256,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,512,0.010007999837398529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,64,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,128,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,128,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,32,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,256,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,2560,0.013812444276279874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,65536,0.02753777801990509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,64,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,51200,0.02183644473552704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,128,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,32,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,65536,0.054173333777321704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,65536,0.047127998537487455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,16384,0.010773333410422007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,51200,0.04376799861590067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,51200,0.03763733307520548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,65536,0.10904355843861897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,16384,0.015523556205961438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,16384,0.015849777393870883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,10240,0.007956444389290279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,51200,0.08750666512383355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,16384,0.03468799922201369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,12288,0.013196444345845116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,12288,0.012392000191741519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,8192,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,10240,0.011737777955002256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,10240,0.011146666275130378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,12288,0.028780443800820246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,2560,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,7168,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,8192,0.010598222414652506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,6144,0.007532444265153673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,10240,0.025454221500290766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,7168,0.010017777482668558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,8192,0.021976000732845728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,7168,0.009373333719041612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,5120,0.00630044440428416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,6144,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,6144,0.008734222087595198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,4096,0.006721777634488211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,5120,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,5120,0.008264888491895463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,6144,0.0189057770702574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,12288,0.00850399997499254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,3584,0.006072000082996156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,4096,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,4096,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,3072,0.005320888840489917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,3584,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,3584,0.006996444529957241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,4096,0.016207110550668504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,2560,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,3072,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,3072,0.007243555453088548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,3584,0.015119110544522604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,8192,0.009659555223253038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,2560,0.007002666592597961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,2560,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,2048,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,3072,0.014779556128713818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,1536,0.0047048889100551605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,2048,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,2560,0.013453333742088742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,2048,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,7168,0.020226667324701946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,1024,0.0038488888078265717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,1536,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,2048,0.01313244468635983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,768,0.0032311110860771606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,1024,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,5120,0.01686755485004849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,1536,0.011772444678677453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,1024,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,512,0.0028479999552170434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,768,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,768,0.005753777921199799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,256,0.0022435554613669715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,512,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,512,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,768,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,128,0.0021928888228204516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,512,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,256,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,256,0.004949333353175057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,64,0.0018746666610240936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,128,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,32,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,256,0.010382222632567087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,65536,0.021780444516075984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,64,0.004949333353175057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,128,0.007948444121413762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,1536,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,51200,0.017619555195172627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,32,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,65536,0.05407199925846524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,65536,0.04660444458325704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,16384,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,51200,0.04344177908367581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,51200,0.037316444847318865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,1024,0.010911111202504901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,12288,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,65536,0.10895644293891059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,16384,0.014696000350846184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,16384,0.015819556183285184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,10240,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,51200,0.08653510941399468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,12288,0.013094222380055321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,16384,0.034288889831966825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,12288,0.012118221984969245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,10240,0.011720889144473605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,10240,0.011028444601429833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,7168,0.006779555645253923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,12288,0.028537776735093858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,8192,0.010688888529936472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,8192,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,10240,0.02572888963752323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,6144,0.006189333481921091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,7168,0.010013333625263637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,5120,0.0063591111037466265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,7168,0.009311111436949836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,8192,0.021931555536058214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,6144,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,4096,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,7168,0.02028622229894002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,5120,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,5120,0.008078222473462423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,6144,0.019309333629078336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,3584,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,4096,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,4096,0.007300444775157505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,5120,0.01696711116366916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,3072,0.005513777749405966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,3584,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,3584,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,2560,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,4096,0.015917332635985482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,3072,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,8192,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,3584,0.015250666273964776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,3072,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,2560,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,2560,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,3072,0.013809777796268463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,2048,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,2048,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,2560,0.013772444592581855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,1024,0.003535110917356279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,6144,0.00904888908068339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,1536,0.005717333406209946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,2048,0.012447111308574677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,768,0.0031386667655573953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,1024,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,1024,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,1536,0.011499555574523078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,768,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,768,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,512,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,768,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,512,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,2048,0.004217777815130022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,256,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,512,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,512,0.010424888796276517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,128,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,256,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,256,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,128,0.005466666486528184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,1536,0.0063742221229606206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,128,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,64,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,32,0.001800000046690305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,65536,0.015976000163290236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,128,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,64,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,32,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,51200,0.013090666797426013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,1024,0.01075644459989336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,65536,0.0454364452097151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,65536,0.05347999930381775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,16384,0.007825777762466008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,51200,0.03833511140611436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,51200,0.03683022326893277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,65536,0.10769155290391709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,12288,0.006849777781301075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,16384,0.014499555031458536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,16384,0.015655999382336933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,768,1536,0.004557333472702238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,51200,0.08682577477561103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,10240,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,12288,0.013182222015327878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,12288,0.012091555529170565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,16384,0.03422133458985223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,10240,0.011739555332395764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,256,0.009683555199040307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,7168,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,12288,0.027775999572541978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,8192,0.010485333700974783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,8192,0.009658666948477427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,10240,0.024680889315075342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,6144,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,7168,0.009747555686367882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,8192,0.021993777818149988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,7168,0.009323555562231276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,5120,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,6144,0.009344889058007134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,6144,0.008634666601816813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,7168,0.020231111182106864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,5120,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,6144,0.01854311095343696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,5120,0.008100444244013892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,3584,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,5120,0.01722577710946401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,4096,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,3584,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,3584,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,4096,0.015830222103330824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,3072,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,8192,0.005680888891220093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,2560,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,10240,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,3072,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,3584,0.015100444356600443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,2048,0.004867555366622077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,3072,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,2560,0.0070426662762959795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,2560,0.006650666809744305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,1536,0.0035502223504914176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,3072,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,2048,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,2048,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,2560,0.013803555733627744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,1024,0.0028568889117903183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,4096,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,1536,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,768,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,2048,0.012554667062229581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,1024,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,1024,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,512,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,4096,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,768,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,768,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,256,0.0022053333620230355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,512,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,512,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,768,0.010792888700962067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,128,0.001964444501532449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,256,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,256,0.005629333357016246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,512,0.009735999835862054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,64,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,128,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,128,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,512,32,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,256,0.009312888814343346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,65536,0.010552888943089379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,64,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,128,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,1536,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,32,0.004600888739029567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,51200,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,65536,0.043860445419947304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,65536,0.04646933409902784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,1536,0.01201333353916804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,51200,0.037245333194732666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,51200,0.03457333313094245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,65536,0.10752533541785346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,1024,0.011403555671374003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,12288,0.005231110999981563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,16384,0.014521777629852295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,51200,0.08600444263882107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,16384,0.015818667080667283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,10240,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,12288,0.01292799992693795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,12288,0.01203733351495531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,8192,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,16384,0.03386844529045953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,10240,0.011728888584507836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,10240,0.011223110887739392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,7168,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,8192,0.010429333481523724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,7168,0.009303111169073317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,8192,0.02164177762137519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,6144,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,16384,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,7168,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,7168,0.020455999506844413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,5120,0.0052986666560173035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,6144,0.008643555144468943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,5120,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,6144,0.019283556275897555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,4096,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,5120,0.008571555217107138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,5120,0.016827555166350473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,12288,0.02774577836195628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,3584,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,4096,0.007927999728255803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,10240,0.024724443753560383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,3584,0.00737955586777793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,4096,0.015827554795477126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,3584,0.014831110835075378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,3072,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,6144,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,2560,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,3072,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,2560,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,2560,0.006308444258239534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,3072,0.013757333159446716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,2560,0.013431111143694984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,4096,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,2048,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,1536,0.00349866681628757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,2048,0.01239911135700014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,3584,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,3072,0.004567111117972268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,1536,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,1536,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,1536,0.01239466667175293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,1024,0.005980444451173146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,1024,0.0052888890107472735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,768,0.002712888850106133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,768,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,1024,0.01108000013563368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,512,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,768,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,2048,0.004576888763242298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,256,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,768,0.010386666489972008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,512,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,512,0.004944889081848992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,128,0.001818666648533609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,256,0.005335111171007156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,2048,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,8192,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,512,0.010679999987284342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,64,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,32,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,128,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,128,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,65536,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,256,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,64,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,51200,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,32,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,128,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,65536,0.04589866598447164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,16384,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,65536,0.042288889487584434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,51200,0.03712800145149231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,51200,0.03422310948371887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,12288,0.005203555441564984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,65536,0.10698666837480332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,16384,0.014120888378885059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,16384,0.015557333827018738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,51200,0.08579999870724148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,10240,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,12288,0.013163555827405719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,12288,0.012099555797047086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,256,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,12288,0.027515555421511333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,10240,0.010933332973056369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,8192,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,10240,0.02439111140039232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,256,1024,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,7168,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,8192,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,8192,0.021939555803934734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,7168,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,6144,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,7168,0.019725332657496136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,6144,0.008640000389681922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,6144,0.009029332962301042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,5120,0.004229333251714706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,16384,0.03398666779200236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,6144,0.018872888551818002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,4096,0.004404444247484207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,5120,0.008366222182909647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,5120,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,4096,0.007325333025720384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,4096,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,3584,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,4096,0.015825778245925903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,3584,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,3072,0.004598222259018156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,7168,0.009030222064918941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,3584,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,3584,0.014814222852389017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,3072,0.007358222372002072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,2560,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,3072,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,2560,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,3072,0.013807999591032663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,2560,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,2048,0.003909333298603694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,2560,0.01368444412946701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,10240,0.011824000212881299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,1536,0.00350577798154619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,2048,0.00612088872326745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,1024,0.002842666581273079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,5120,0.017148445049921673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,1536,0.006261333409282897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,8192,0.01071200023094813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,768,0.0025617778301239014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,1536,0.0058853332367208265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,1024,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,1024,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,1536,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,512,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,768,0.0058400001790788435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,768,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,256,0.0021217777911159727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,1024,0.01038755559258991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,512,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,768,0.011382222175598145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,128,0.0018071112119489247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,256,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,256,0.004948444250557157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,512,0.0099982221921285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,64,0.001812444378932317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,128,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,128,32,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,128,0.004959110998445087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,65536,0.007055111229419708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,2048,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,51200,0.006811555474996567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,64,0.004856889032655292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,16384,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,128,0.0064222220745351576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,12288,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,32,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,10240,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,2048,0.01277511070171992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,65536,0.04214488797717624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,8192,0.004983110974232356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,51200,0.033610665135913424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,7168,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,16384,0.014333332578341166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,6144,0.004755555755562252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,5120,0.0042168887125121225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,12288,0.011764444410800934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,4096,0.003910222401221593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,10240,0.011034666664070554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,8192,0.00999644481473499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,3584,0.004621333132187526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,7168,0.009033777647548253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,6144,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,2560,0.00453955555955569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,512,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,2048,0.004466666529575984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,4096,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,1536,0.0032026666320032547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,3584,0.007249777515729268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,1024,0.0028924445311228433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,3072,0.00665866666369968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,2560,0.006313777632183499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,768,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,2048,0.006080889039569431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,512,0.002471999989615546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,1536,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,256,0.001806222316291597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,1024,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,128,0.0018079999006456798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,768,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,64,0.0018079999006456798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,512,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,32,0.0018035556293196147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,128,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,65536,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,64,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,51200,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,32,0.004872888740566042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,16384,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,65536,0.042100446091757886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,5120,0.007761778102980719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,12288,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,10240,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,51200,0.03360177742110358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,8192,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,16384,0.01426755554146237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,7168,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,12288,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,256,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,6144,0.004557333472702238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,10240,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,5120,0.004619555754794015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,4096,0.0038782221575578055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,8192,0.009665778113736046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,3584,0.004305777864323722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,7168,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,3072,0.004228444563017951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,6144,0.008352888955010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,5120,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,2560,0.0041804446114434134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,4096,0.007145778172545963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,256,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,2048,0.0041244443919923575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,1536,0.0032231110251612137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,3584,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,1024,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,3072,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,768,0.002479111154874166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,512,0.0021359999146726397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,2560,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,256,0.0021520000365045336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,2048,0.006169777777459886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,1536,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,128,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,64,3072,0.004600000050332812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,1024,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,64,0.001655111089348793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,12,32,32,0.0016151110952099164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,768,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,512,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,256,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,64,0.004947555561860402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,32,0.004610666798220741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,16384,0.32231465975443524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,16384,0.16277778148651123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,51200,0.48808887269761825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,12288,0.22567645708719888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,16384,0.16018577416737875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,51200,0.960784912109375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,12288,0.12125244405534531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,8192,0.15355822775099012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,10240,0.18752177556355795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,12288,0.10304177469677395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,10240,0.15309955014122859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,10240,0.08785866366492377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,7168,0.13231821854909262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,128,0.004578666554556953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,6144,0.11424088478088379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,51200,0.43607645564609104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,7168,0.06276622083452013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,5120,0.0959920022222731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,6144,0.05480622251828512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,8192,0.07146577702628241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,5120,0.04684177703327603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,4096,0.08016977707544963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,3584,0.067994667424096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,4096,0.03899022274547153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,3072,0.05926489167743259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,4096,0.06590311394797431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,3584,0.05295377638604906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,3584,0.03489333391189575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,2560,0.04987555411126879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,3072,0.031522668070263334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,3072,0.04697688751750522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,2048,0.040848887628979154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,2560,0.026983110441101923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,2560,0.040630221366882324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,1536,0.031530665026770696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,7168,0.0996933314535353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,2048,0.0284551117155287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,2048,0.03464711043569777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,1024,0.022255111071798537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,1536,0.01908622185389201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,1536,0.02384177843729655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,1024,0.01548266741964552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,768,0.017723555366198223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,512,0.012618666721714867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,1024,0.020939555433061387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,5120,0.07245866457621257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,256,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,768,0.014813333749771118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,512,0.012836444709036084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,768,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,128,0.005176888985766305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,512,0.01276622215906779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,64,0.004211555752489301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,256,0.00998933364947637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,128,0.010497777826256223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,128,0.008027555214034187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,64,0.009397333694828881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,32,0.009942221972677443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,8192,0.11104800303777058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,6144,0.08151466978920831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,65536,0.5144337548149956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,256,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,65536,1.086846245659722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,51200,0.8478710916307238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,32,0.0041884444653987885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,51200,0.3902017805311415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,16384,0.2780746618906657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,12288,0.20733243889278838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,10240,0.17368355062272814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,16384,0.13075466950734457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,12288,0.09781689114040798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,65536,0.4513004620869954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,12288,0.08538933595021565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,10240,0.07122311327192518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,16384,0.12203378147549099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,8192,0.1407324473063151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,10240,0.1252044439315796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,7168,0.12212355931599934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,51200,0.34348000420464414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,8192,0.08929777807659572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,8192,0.0596275528271993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,5120,0.08885155783759223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,7168,0.05013155606057909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,6144,0.04576711191071404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,7168,0.07502222061157227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,6144,0.06936799817615084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,3584,0.06278222137027316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,5120,0.03933866818745931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,5120,0.057887997892167836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,3072,0.0547928876347012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,4096,0.05275377631187439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,3584,0.02912799848450555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,3584,0.04620711008707682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,2560,0.046127110719680786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,2048,0.037236445479922824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,3072,0.02648799949222141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,3072,0.037992000579833984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,2560,0.02290311124589708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,1536,0.02895555562443203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,2560,0.034260445170932345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,2048,0.024567110670937434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,1024,0.02037244372897678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,2048,0.028396444188223943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,1536,0.017871111631393433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,768,0.015107555521859063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,1536,0.019855999284320407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,512,0.0107004443804423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,768,0.012679110798570844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,6144,0.10514222251044379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,256,0.006367110957702001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,768,0.01261333293384976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,512,0.010712888505723743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,128,0.0034808889031410217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,4096,0.073381331231859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,256,0.009368889033794403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,64,0.0028088889602157804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,4096,0.03277066681120131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,128,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,128,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,64,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,256,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,32,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,65536,0.3701475461324056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,51200,0.2702515655093723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,65536,0.17546044455634224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,1024,0.017496888836224873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,16384,0.08951377868652344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,65536,0.15855644808875188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,51200,0.13121867179870605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,512,0.011067555182509951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,12288,0.06929333342446221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,10240,0.05182844400405884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,51200,0.1261048846774631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,16384,0.05137155453364054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,32,0.0028364445186323593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,16384,0.048001776138941445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,8192,0.0436426666047838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,7168,0.03769244419203864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,10240,0.04090933336151971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,10240,0.029269334342744615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,8192,0.02479377720091078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,8192,0.03733155462476943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,6144,0.032101333141326904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,7168,0.022279999322361414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,5120,0.028188443846172754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,6144,0.020045333438449435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,1024,0.013456888496875763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,6144,0.028204444381925795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,4096,0.02405511173937056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,5120,0.018029333816634286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,3584,0.020040000478426617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,5120,0.023281777898470562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,4096,0.015488000379668342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,3072,0.017808889349301655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,4096,0.021371554997232225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,3584,0.014440000057220459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,2560,0.015101333459218344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,3072,0.015388444066047668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,12288,0.033195555210113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,12288,0.036355555057525635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,2048,0.0124177775449223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,2560,0.014711999230914645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,3072,0.015969778100649517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,1536,0.009955555200576782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,2560,0.015187554889255099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,2048,0.011028444601429833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,1536,0.009519110951158736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,2048,0.01331733332739936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,768,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,1536,0.01054133309258355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,1024,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,7168,0.031255112753974065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,512,0.0041928887367248535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,1024,0.009328000247478485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,256,0.002499555548032125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,768,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,512,0.00723555518521203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,128,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,512,0.006715555571847492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,256,0.006656000183688269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,256,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,128,0.006700444552633498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,32,0.0021582223061058256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,128,0.00553688903649648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,3584,0.01818933255142636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,64,0.005893333504597346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,65536,0.2517351044548882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,1024,0.007615110940403408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,65536,0.1455706622865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,51200,0.19794755511813691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,65536,0.13981689347161189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,768,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,16384,0.06503466765085857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,51200,0.11226133505503337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,51200,0.11026489072375828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,16384,0.04321777820587158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,16384,0.04125600059827169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,10240,0.04310577776696947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,8192,0.03568889035118951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,12288,0.030049777693218652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,10240,0.025763556361198425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,7168,0.03118577930662367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,10240,0.03755911191304525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,8192,0.021977777282396953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,6144,0.027463111612531874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,8192,0.03075288732846578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,7168,0.020004444652133517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,32,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,7168,0.027628444963031348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,5120,0.023780445257822674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,6144,0.018228444788191054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,4096,0.019140443868107267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,6144,0.02384800049993727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,5120,0.016128000285890367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,3584,0.01661777827474806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,5120,0.020911999874644812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,4096,0.014335999886194864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,3072,0.014542222023010254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,4096,0.019274666905403137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,3584,0.013443555268976422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,2560,0.012362666428089142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,3072,0.012392000191741519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,12288,0.05059822069274055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,2048,0.010344888601038191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,3072,0.014827556080288358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,2560,0.01149777736928728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,1536,0.00850577818022834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,2560,0.014121777481502958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,64,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,2048,0.009892444643709395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,12288,0.031762666172451444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,1024,0.006882666713661618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,768,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,2048,0.012458666331238218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,1536,0.008957333034939235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,1024,0.007456888755162557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,1536,0.009866666462686326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,768,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,256,0.0028328889360030494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,1024,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,768,0.006828444285525217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,512,0.006919999917348226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,128,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,512,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,256,0.006497777584526274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,64,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,32,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,256,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,128,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,64,0.005773333211739858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,65536,0.21578578154246011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,32,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,3584,0.01796533332930671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,65536,0.14177155494689941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,51200,0.17110044426388213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,16384,0.057909336354997426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,65536,0.12647910912831625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,12288,0.04495111107826233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,51200,0.10727110836240981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,512,0.0038151111867692736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,51200,0.10037510924869114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,16384,0.039577778842714094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,16384,0.039800001515282526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,10240,0.03676977753639221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,8192,0.030768000417285498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,12288,0.028354667954974707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,12288,0.030449777841567993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,7168,0.027698665857315063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,8192,0.02085599965519375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,8192,0.026718222432666357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,128,0.006694222076071634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,7168,0.019088000059127808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,7168,0.023318222827381555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,5120,0.02100444502300686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,6144,0.017276444368892245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,4096,0.017695110705163743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,5120,0.015538665983412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,6144,0.01978933314482371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,3584,0.015205333630243937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,5120,0.017880888448821176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,4096,0.013645333548386892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,3072,0.013384888569513956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,4096,0.01611911091539595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,3584,0.012892444100644855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,2560,0.011389333340856763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,3584,0.014660444524553088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,2048,0.009816889133718278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,3072,0.01260355528857973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,2560,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,1536,0.008032889001899296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,2560,0.012420444852775998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,2048,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,1024,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,2048,0.010971555279360877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,10240,0.025006221400366888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,1536,0.008266666697131263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,10240,0.03157511022355821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,1536,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,1024,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,512,0.0038968887594011095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,1024,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,768,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,256,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,512,0.006758222149478064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,768,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,128,0.0025031111306614345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,256,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,64,0.0021297778520319196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,256,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,128,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,32,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,128,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,64,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,65536,0.17241422335306802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,32,0.005719111197524601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,51200,0.12850133577982584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,65536,0.07107466459274292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,65536,0.09071733554204305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,16384,0.044136888451046415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,768,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,6144,0.02442933287885454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,51200,0.057320886188083224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,16384,0.024489776955710515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,16384,0.0274000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,10240,0.026567111412684124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,12288,0.018532445033391316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,8192,0.02405511173937056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,12288,0.021613332960340712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,10240,0.016473778420024447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,7168,0.02085777786042955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,10240,0.023082666926913794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,8192,0.014655111564530266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,6144,0.017996443642510306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,8192,0.02090222140153249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,7168,0.013639111485746173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,3072,0.011615110768212212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,7168,0.018394667241308425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,5120,0.015468445089128284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,6144,0.012448888685968188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,4096,0.013421333498424955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,6144,0.016188444362746347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,5120,0.011648000114493899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,3584,0.011683555940786997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,5120,0.014647111296653748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,4096,0.010547555155224271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,51200,0.0690773328145345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,3072,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,4096,0.013082666529549493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,3584,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,2560,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,3584,0.01238044434123569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,2048,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,3072,0.01071466671095954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,2560,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,1536,0.006607111129495833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,2560,0.010399999717871347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,512,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,1024,0.0047004446387290955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,2048,0.009295110901196798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,1536,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,1536,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,1024,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,1024,0.00665511108107037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,512,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,768,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,768,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,256,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,512,0.005727111051479976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,256,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,256,0.005889777921968036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,128,0.0021564445147911706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,3072,0.008714666797055138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,64,0.0019662222928471034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,128,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,32,0.002143111079931259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,128,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,2048,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,64,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,65536,0.14813689390818277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,32,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,768,0.004008000095685323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,51200,0.11536799536810981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,16384,0.04023377762900458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,65536,0.08336622185177273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,12288,0.03163733416133457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,51200,0.05367288986841837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,16384,0.021742223037613764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,51200,0.06260533465279473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,512,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,12288,0.01770755483044518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,12288,0.019902222686343722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,10240,0.01568800045384301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,10240,0.021922666165563796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,12288,0.034648001194000244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,7168,0.0196088883611891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,8192,0.013759999639458127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,6144,0.01686755485004849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,8192,0.01956355571746826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,7168,0.012999110751681857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,5120,0.014722666806644864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,7168,0.016629333297411602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,6144,0.01551555593808492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,4096,0.012798222402731577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,5120,0.011010666688283285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,5120,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,3584,0.011039999624093374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,4096,0.009683555199040307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,65536,0.0656871133380466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,4096,0.01273333364062839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,3072,0.009738666315873465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,3584,0.009190222455395592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,16384,0.02568000058333079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,2560,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,3072,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,3584,0.011511999699804517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,2048,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,3072,0.010288000106811523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,8192,0.0229759994480345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,1536,0.006220444622966979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,2560,0.007932444413503012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,2560,0.009875555833180746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,1024,0.00452622233165635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,2048,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,1536,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,768,0.003597333199448056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,2048,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,512,0.0028479999552170434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,1024,0.006350222147173352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,768,0.005977777971161737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,1024,0.006313777632183499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,256,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,6144,0.012071111136012606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,512,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,768,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,128,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,64,0.0019822222077184254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,512,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,256,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,32,0.0019991110182470745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,128,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,256,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,128,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,65536,0.1211128897137112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,64,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,32,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,51200,0.09569511148664687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,10240,0.026612444056404963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,16384,0.03390222125583225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,65536,0.07436088720957439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,65536,0.0640960004594591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,12288,0.027051554785834417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,51200,0.056589331891801625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,51200,0.05152711272239685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,16384,0.02068533334467146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,10240,0.02304977840847439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,12288,0.01720977822939555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,16384,0.023567110300064087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,12288,0.017954667409261067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,10240,0.015254222684436373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,7168,0.017530666457282174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,8192,0.013425778183672162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,10240,0.020615110794703167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,8192,0.018554665976100497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,7168,0.01272266689274046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,5120,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,7168,0.016138666205936007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,6144,0.011758222348160215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,4096,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,6144,0.014984000060293408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,5120,0.010664000279373594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,3584,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,5120,0.013269333375824822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,4096,0.008750222623348236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,3072,0.009148444566461775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,4096,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,3584,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,2560,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,3584,0.011277333729796939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,3072,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,2048,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,3072,0.009735999835862054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,2560,0.007301333049933116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,1536,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,2560,0.009311999711725447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,8192,0.019864888654814828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,1024,0.003927999900446998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,2048,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,2048,0.007668444679843054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,1536,0.006903111106819577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,1536,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,6144,0.01556799974706438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,1536,0.006717333363162146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,512,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,1024,0.005998222364319696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,256,0.0024604443460702896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,768,0.006187555690606435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,512,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,128,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,256,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,512,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,64,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,256,0.005199111170238919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,32,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,128,0.005630222045713001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,128,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,65536,0.1053804424073961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,64,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,32,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,51200,0.08400089210934109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,65536,0.06655910942289564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,65536,0.06134577592213949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,16384,0.030470222234725952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,51200,0.04948000113169352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,12288,0.024767110745112102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,51200,0.05130577749676175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,768,0.0032631111227803757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,16384,0.01997599999109904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,10240,0.02068533334467146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,16384,0.02161688937081231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,1024,0.006265777680608962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,12288,0.016657777958446078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,10240,0.015032889114485847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,768,0.005881777654091517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,7168,0.016001777516471017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,10240,0.018681777848137748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,6144,0.014167999227841696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,8192,0.017092444830470614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,7168,0.01219555570019616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,5120,0.012235555383894177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,6144,0.011379555695586734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,7168,0.015871110889646742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,4096,0.011002666420406766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,6144,0.01372977760103014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,3584,0.009562667045328353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,5120,0.009869332942697737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,4096,0.008353778057628209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,5120,0.01276000009642707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,3072,0.008340444829728868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,3584,0.008072000410821702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,4096,0.012094222009181976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,2560,0.0075297777851422625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,3584,0.01036888857682546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,3072,0.007619555625650618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,2048,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,2560,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,2560,0.00871288859181934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,1536,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,2048,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,8192,0.018211555149820115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,2048,0.007618666523032718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,1024,0.00369422220521503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,12288,0.016568000117937725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,1536,0.006598222172922558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,768,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,1536,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,8192,0.013108444710572561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,512,0.0028497777465316984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,768,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,1024,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,256,0.0021928888228204516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,512,0.005670222143332164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,768,0.005643555687533484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,128,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,512,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,64,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,128,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,256,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,32,0.001956444440616502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,128,0.004965333475006951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,65536,0.085116446018219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,64,0.005092444519201915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,32,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,3072,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,51200,0.06619111034605238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,65536,0.060432891050974526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,16384,0.025079111258188885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,65536,0.05884710947672526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,51200,0.047535998953713306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,51200,0.04669244421852959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,16384,0.019498666127522785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,16384,0.019331556227472093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,10240,0.017609778377744887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,1024,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,12288,0.01612444387541877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,8192,0.01551555593808492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,12288,0.015453333655993143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,7168,0.012834666503800286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,10240,0.014466666513019137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,8192,0.01276622215906779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,10240,0.01772622267405192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,6144,0.011343999869293638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,256,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,8192,0.01608088943693373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,7168,0.011731555892361535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,5120,0.00998133338159985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,4096,0.008956444760163626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,7168,0.0140328887436125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,6144,0.01034222212102678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,3584,0.00776711106300354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,5120,0.008775110873911116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,6144,0.012880889077981314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,3072,0.007156444092591603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,5120,0.012058667010731168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,4096,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,2560,0.006213333457708359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,4096,0.011370666325092316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,3584,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,2048,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,3072,0.007618666523032718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,2560,0.007085333267847697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,3072,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,1536,0.004042666819360522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,2048,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,2560,0.007795555724038019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,1024,0.0031795555518733132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,2048,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,1536,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,1536,0.0064444442590077715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,1024,0.006146666490369373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,512,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,768,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,1024,0.0058755555914507965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,256,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,768,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,512,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,128,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,256,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,512,0.005392888767851724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,64,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,32,0.001962666710217794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,256,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,128,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,128,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,65536,0.07675821913613214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,64,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,51200,0.058837334314982094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,32,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,16384,0.022075555390781824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,65536,0.058092441823747426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,65536,0.0575431121720208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,3584,0.009299555586444007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,12288,0.0177502218219969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,51200,0.04682133264011807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,51200,0.044920000765058726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,10240,0.015360888507631091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,16384,0.019281778070661757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,16384,0.01856355534659492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,12288,0.015825778245925903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,768,0.0028817777832349143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,8192,0.01314222233162986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,12288,0.014783999986118741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,10240,0.01718488832314809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,8192,0.014416888356208801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,6144,0.010701333483060202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,7168,0.013811555173661975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,6144,0.010001777774757808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,5120,0.00963644435008367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,6144,0.011572444604502784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,5120,0.008958222137557136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,4096,0.008653333617581261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,5120,0.011419555379284753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,12288,0.02015733387735155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,4096,0.008355555435021719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,3584,0.0074622225430276655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,4096,0.008997333546479544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,3072,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,7168,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,3584,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,2560,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,8192,0.0123831108212471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,3584,0.008347555167145198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,2048,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,3072,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,7168,0.011076444553004371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,1536,0.00387999994887246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,2560,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,2048,0.0069164443347189165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,1024,0.0032053333189752367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,2048,0.006664000037643645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,1536,0.006549333532651265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,768,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,1536,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,1024,0.00602311102880372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,512,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,768,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,1024,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,256,0.0021946666141351065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,768,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,512,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,128,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,256,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,512,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,64,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,32,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,128,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,128,0.004990222139490975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,10240,0.014171555638313293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,64,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,32,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,51200,0.048912889427608915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,3072,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,16384,0.01921422282854716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,65536,0.055408888392978244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,65536,0.05710222323735555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,2560,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,12288,0.015578667322794596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,51200,0.0458808872434828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,10240,0.01514311134815216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,51200,0.043206221527523465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,16384,0.018928888771269057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,8192,0.013356444736321768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,12288,0.015544000599119397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,16384,0.017553778158293832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,7168,0.012096888489193387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,12288,0.014472888575659858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,10240,0.016001777516471017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,6144,0.010823999842007955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,8192,0.013502221968438892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,7168,0.01220266686545478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,5120,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,6144,0.009680888719028896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,256,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,6144,0.01180355581972334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,65536,0.061591108640034996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,5120,0.008847999903890822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,3584,0.007230222225189209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,5120,0.009853333234786987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,4096,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,3072,0.006568888823191325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,4096,0.008821333448092142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,3584,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,2560,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,3584,0.007724444071451823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,3072,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,2048,0.004576888763242298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,3072,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,2560,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,1536,0.0038533334930737815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,2560,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,10240,0.013920888304710388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,2048,0.0069102222720781965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,1024,0.0031475555151700974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,8192,0.011691555380821228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,1536,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,7168,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,768,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,1536,0.006149333384301927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,1024,0.006233777850866318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,512,0.00249777775671747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,1024,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,768,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,256,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,4096,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,512,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,768,0.0053022222386466126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,128,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,64,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,256,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,512,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,32,0.0017973333597183228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,128,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,256,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,65536,0.05567110909356011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,128,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,51200,0.04345688886112637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,64,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,16384,0.019721777902709115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,32,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,65536,0.056178669134775795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,65536,0.05261422197024027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,51200,0.04520977867974175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,51200,0.041689776711993747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,16384,0.018788443671332467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,10240,0.013910222384664746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,16384,0.01721155477894677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,12288,0.015174221661355762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,8192,0.01275288893116845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,2048,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,12288,0.014121777481502958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,10240,0.013448889056841532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,8192,0.011028444601429833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,10240,0.014412444498803882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,6144,0.010231110784742568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,7168,0.010024000373151567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,5120,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,7168,0.012097777591811286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,6144,0.00943999985853831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,4096,0.007909333540333642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,6144,0.010390222072601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,5120,0.00868622213602066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,5120,0.008785777621799046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,3584,0.007078222102589077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,4096,0.008018666671382057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,3072,0.00620888877246115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,3584,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,3584,0.00739822205570009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,2560,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,12288,0.016306666864289176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,3072,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,2048,0.004247111164861255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,2560,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,2560,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,7168,0.011113777756690979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,1536,0.0038702223036024305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,2048,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,1024,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,2048,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,1536,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,8192,0.012321777641773224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,768,0.0028524444335036804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,512,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,1536,0.005975111077229182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,1024,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,1024,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,768,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,512,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,768,0.005407111098368962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,256,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,256,0.004953777624501122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,4096,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,64,0.0017973333597183228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,128,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,32,0.0017973333597183228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,128,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,65536,0.04841599861780802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,3072,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,51200,0.03527644607755873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,64,0.004871110949251386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,65536,0.055162668228149414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,65536,0.05024266574117872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,51200,0.04475288920932346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,51200,0.040227555566363864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,12288,0.01312444441848331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,16384,0.018252443936136033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,16384,0.016725333200560678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,10240,0.011914666328165265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,256,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,12288,0.013530666629473368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,128,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,10240,0.0123831108212471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,8192,0.010994666980372535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,10240,0.013215111361609565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,7168,0.00965244405799442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,8192,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,8192,0.012252444194422828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,7168,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,6144,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,7168,0.01016444464524587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,5120,0.008040000167157915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,32,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,6144,0.009383999639087254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,4096,0.007239110767841339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,6144,0.009032888544930352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,5120,0.009033777647548253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,3584,0.006559111177921295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,5120,0.008354666332403818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,4096,0.008006222546100616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,3072,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,4096,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,3584,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,2560,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,12288,0.013783111340469785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,3584,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,2048,0.0042275554604000514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,3072,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,3072,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,2560,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,1536,0.003702222059170405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,2560,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,1024,0.00319377767542998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,2048,0.006166222194830577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,1536,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,768,0.002874666617976295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,1024,0.006211555666393704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,1024,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,512,0.002519999941190084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,768,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,16384,0.01604977746804555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,768,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,256,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,512,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,512,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,128,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,256,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,64,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,256,0.004951111144489712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,32,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,65536,0.034689777427249484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,128,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,64,0.005105777747101254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,512,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,51200,0.027403554982609216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,32,0.004872888740566042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,16384,0.012631999949614206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,2048,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,65536,0.054660442802641124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,12288,0.01072177787621816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,65536,0.04864355590608385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,1536,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,10240,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,51200,0.044201778040991895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,16384,0.017176000608338248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,8192,0.008981333010726506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,51200,0.03867200016975403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,12288,0.013635555903116861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,16384,0.01629066632853614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,7168,0.008351999852392409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,6144,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,12288,0.013312000367376538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,8192,0.010680889089902243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,10240,0.012444444828563266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,5120,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,7168,0.010339555641015371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,8192,0.00999466660949919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,4096,0.00701422244310379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,6144,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,7168,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,3584,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,5120,0.008632889224423302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,3072,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,4096,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,5120,0.008340444829728868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,3584,0.007666666474607255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,2560,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,4096,0.007345777418878343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,2048,0.00424177779091729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,3072,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,3584,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,1536,0.0035546666218174826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,2560,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,3072,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,1024,0.003176888864901331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,2048,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,2560,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,768,0.0028604444944196274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,2048,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,1024,0.006157333238257303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,1536,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,512,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,1024,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,768,0.005315555466545953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,256,0.0021582223061058256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,512,0.005790222022268508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,10240,0.012071111136012606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,128,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,512,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,256,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,64,0.0018053334206342697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,256,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,32,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,128,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,6144,0.008748444418112436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,65536,0.027992000182469685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,128,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,51200,0.02200711104604933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,64,0.004960888789759742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,32,0.005075555708673266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,65536,0.054487999942567616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,16384,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,65536,0.04691377613279554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,51200,0.043771555026372276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,12288,0.00851733320289188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,51200,0.03767199979888068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,16384,0.015836444165971544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,1536,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,12288,0.013065777719020844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,16384,0.015455111861228943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,8192,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,12288,0.012296000288592445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,10240,0.011750222080283694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,7168,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,10240,0.011396444506115384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,8192,0.010508444574144153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,6144,0.007341333561473423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,7168,0.010045333041085137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,8192,0.009918221996890174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,5120,0.006303999986913469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,7168,0.009301332963837517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,6144,0.009354666703277165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,4096,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,6144,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,5120,0.008681777450773451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,3584,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,4096,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,5120,0.008207999997668797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,3072,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,4096,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,3584,0.007677333222495184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,2560,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,3072,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,2048,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,2560,0.007010666860474481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,3072,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,1536,0.004980444494220945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,2560,0.00666133314371109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,2048,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,1024,0.003535110917356279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,2048,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,1536,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,768,0.003489777859714296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,768,0.005990222096443176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,1536,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,1024,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,1024,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,768,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,256,0.002191999927163124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,768,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,512,0.005631999837027655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,128,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,512,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,256,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,64,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,256,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,128,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,32,0.001964444501532449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,65536,0.021689777572949726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,128,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,51200,0.017525333497259352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,3584,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,64,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,16384,0.009109333157539368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,65536,0.053852445549435086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,32,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,12288,0.00792711145348019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,65536,0.04603644543223911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,51200,0.043460445271597974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,10240,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,51200,0.03733777668741014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,12288,0.013139555851618448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,16384,0.014730667074521383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,8192,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,12288,0.012223111258612739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,10240,0.011789333489206104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,7168,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,512,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,10240,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,8192,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,7168,0.00998399986161126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,8192,0.009703999592198266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,5120,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,7168,0.009364444348547194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,6144,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,6144,0.008731555607583787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,4096,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,5120,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,3584,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,5120,0.008031110796663497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,4096,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,3072,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,4096,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,3584,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,2560,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,3584,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,3072,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,2048,0.0041964443193541635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,3072,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,2560,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,16384,0.015841777125994366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,1536,0.004550222307443619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,2560,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,1024,0.0035217776894569397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,1536,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,1536,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,6144,0.005974222388532426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,768,0.0030097777230872046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,1024,0.006051555689838197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,1024,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,512,0.002568888788421949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,768,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,256,0.0022115556316243275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,512,0.0057884446448749965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,768,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,128,0.0018968888454967076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,512,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,256,0.005647111270162795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,64,0.0018800000349680583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,256,0.004949333353175057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,128,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,768,32,0.0018542222678661346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,65536,0.015802666544914246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,128,0.004853333450025982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,51200,0.013056889176368713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,64,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,32,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,10240,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,65536,0.052087999052471586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,16384,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,65536,0.045576890309651695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,12288,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,2048,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,10240,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,51200,0.0366195539633433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,2048,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,12288,0.013071999781661563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,8192,0.005633777628342311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,16384,0.014463111758232117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,7168,0.0052986666560173035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,12288,0.012078222301271228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,10240,0.012039999994966718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,8192,0.010724444356229572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,10240,0.011015111373530494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,7168,0.010054222411579555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,5120,0.0063528890411059065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,6144,0.009319999979601966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,7168,0.009365333451165093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,4096,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,6144,0.00885688927438524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,5120,0.008632889224423302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,3584,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,5120,0.008024888734022776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,4096,0.00811111099190182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,3072,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,4096,0.007155555817815993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,3584,0.007679111427730984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,2560,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,3072,0.0073315559162033936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,3072,0.006650666809744305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,51200,0.03847288754251268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,2048,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,2560,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,1536,0.0036115555299652945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,2560,0.006568888823191325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,2048,0.00685777763525645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,1024,0.0028275555620590844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,2048,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,1536,0.0063164445261160536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,6144,0.005884444548024072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,768,0.0028400001012616684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,1536,0.005656888915432825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,8192,0.009711110757456886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,512,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,1024,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,256,0.0022231110682090125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,768,0.005635555419656966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,1024,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,128,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,512,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,256,0.005427555491526921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,512,0.004952888935804367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,64,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,256,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,512,32,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,128,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,128,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,65536,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,3584,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,64,0.004597333156400257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,51200,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,32,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,65536,0.04625155528386434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,16384,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,65536,0.04282222191492716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,12288,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,51200,0.03731111023161147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,16384,0.015786666009161208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,10240,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,51200,0.03428444597456191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,16384,0.014496889379289416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,12288,0.013068444199032254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,8192,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,12288,0.012095111111799875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,10240,0.011706666813956367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,7168,0.005505777895450592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,8192,0.01038755559258991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,10240,0.01108444482088089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,6144,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,8192,0.00963288876745436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,7168,0.010001777774757808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,5120,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,768,0.005402666827042897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,7168,0.009200000100665623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,4096,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,6144,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,3584,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,5120,0.008789333204428354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,6144,0.00870577742656072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,4096,0.008015111088752747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,5120,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,3584,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,4096,0.007248889240953658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,2560,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,3072,0.00739911115831799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,2048,0.004583999928500918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,2560,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,2560,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,1536,0.0034977777136696708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,16384,0.015857777661747403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,2048,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,1024,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,1536,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,768,0.0028488888508743713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,1536,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,1024,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,512,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,1024,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,768,0.005681777579916849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,256,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,768,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,512,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,128,0.0018115554832749898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,512,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,3072,0.004243555582231945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,64,0.0018213333355055915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,256,0.004544888933499654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,256,32,0.0018159999615616268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,128,0.005306666509972678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,65536,0.007637333538797166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,3584,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,128,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,51200,0.007242666350470648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,3072,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,64,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,65536,0.04609866605864631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,16384,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,32,0.004528000122971005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,12288,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,65536,0.041666666666666664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,51200,0.037031110790040754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,2048,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,10240,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,16384,0.015815110670195687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,8192,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,12288,0.01309511148267322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,16384,0.014255111416180929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,7168,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,10240,0.011711999773979187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,12288,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,6144,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,8192,0.010408889088365767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,10240,0.011048888994587792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,5120,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,7168,0.009970666633711921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,8192,0.00979644474056032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,6144,0.009191999832789103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,5120,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,6144,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,3584,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,4096,0.00794666674402025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,4096,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,3072,0.00461422238085005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,3584,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,3584,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,2560,0.00453333349691497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,3072,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,3072,0.006776000062624614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,2048,0.004549333204825719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,2560,0.0069102222720781965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,2560,0.006319111006127463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,1536,0.003551111039188173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,2048,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,51200,0.03401599989997016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,1024,0.0028373334142896864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,2048,0.006270222365856171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,1536,0.006231110956933763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,768,0.002537777854336633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,1024,0.005981333139869902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,1536,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,512,0.002537777854336633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,4096,0.00452622233165635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,1024,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,256,0.0018542222678661346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,768,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,768,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,5120,0.008058666355080074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,128,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,256,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,512,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,64,0.0018079999006456798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,128,32,0.0018177777528762817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,256,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,128,0.005327111317051782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,65536,0.0073315559162033936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,51200,0.00720888872941335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,128,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,16384,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,64,0.004545777622196409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,12288,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,65536,0.04176622298028734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,10240,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,51200,0.0340008901225196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,16384,0.014464888307783337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,12288,0.011751111182901593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,7168,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,10240,0.01073688848151101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,6144,0.004592000196377437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,8192,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,5120,0.004560889055331548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,7168,0.008967999782827165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,7168,0.009012444151772393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,4096,0.0041662222809261745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,3584,0.004541333350870344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,512,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,6144,0.008464888566070134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,3072,0.00452977791428566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,5120,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,2560,0.004563555535342959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,2048,0.00435288871328036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,4096,0.0069617778062820435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,1536,0.003212444484233856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,3584,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,3072,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,1024,0.0028044444819291434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,768,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,2560,0.006648889018429651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,512,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,32,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,2048,0.006175111151403851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,256,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,128,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,1536,0.005632888939645555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,64,0.0018293333964215384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,1024,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,32,0.0018293333964215384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,64,8192,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,768,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,65536,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,512,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,51200,0.006858666737874349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,256,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,12288,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,128,0.004594666676388847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,64,0.004570666700601578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,10240,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,32,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,8192,0.004767111192146937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,65536,0.04208622376124064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,7168,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,51200,0.033985777033699885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,16384,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,6144,0.0042462220622433555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,5120,0.0042239998777707415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,4096,0.004182222402758068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,12288,0.012079111403889127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,10240,0.010404444403118556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,3584,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,3072,0.004552888787455029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,7168,0.009091555244392818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,6144,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,5120,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,2560,0.004960000101062986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,2048,0.004573333180612988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,4096,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,1536,0.0034808889031410217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,3584,0.006991110742092133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,1024,0.0026968889352348116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,3072,0.006670222100284364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,2560,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,768,0.0025315555847353404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,2048,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,1536,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,256,0.0020088888704776764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,1024,0.005299555758635203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,128,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,768,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,512,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,64,0.0016293333222468693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,32,0.0016133333038952614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,256,0.004861333303981357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,128,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,64,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,32,0.004536889079544279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,8192,0.00941422250535753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,16384,0.3318817880418566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,16384,0.16293688615163168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,51200,0.4884364340040419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,512,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,51200,0.9934035407172309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,12288,0.25305332077874076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,11,32,16384,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,12288,0.1446453332901001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,12288,0.10118044747246636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,16384,0.13614843951331243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,10240,0.18764711750878227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,10240,0.08782577514648438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,8192,0.15368710623847112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,10240,0.14711644914415148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,8192,0.11314045058356391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,51200,0.42711199654473203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,6144,0.11412266890207927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,8192,0.07067733340793185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,5120,0.09593155648973252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,7168,0.0923537810643514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,7168,0.06277066469192505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,6144,0.05522133244408501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,4096,0.07985955476760864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,6144,0.08477689160241021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,5120,0.04714044597413805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,3584,0.06806221935484144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,5120,0.07330666648017035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,3072,0.058993776639302574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,4096,0.039346665143966675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,4096,0.0635146697362264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,2560,0.04965777860747444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,3584,0.03496888942188687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,2048,0.04048177931043837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,3072,0.031039112144046362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,3072,0.04757155643569099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,1536,0.03165244393878513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,2560,0.02699022160636054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,2560,0.037871999873055354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,1024,0.022613333331214056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,2048,0.03428711162673102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,2048,0.029170665476057265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,1536,0.01905066602759891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,1024,0.015142222245534262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,1536,0.026519111461109583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,512,0.012534221841229333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,7168,0.13229333029852972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,256,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,768,0.01623911162217458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,512,0.012805333567990197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,128,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,256,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,256,0.009832888841629028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,128,0.01036711119943195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,64,0.0042204442951414315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,32,0.0042239998777707415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,3584,0.055678221914503306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,64,0.009425777528021071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,32,0.009661333428488838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,768,0.017865777015686035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,1024,0.019490665859646268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,768,0.014815110299322339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,512,0.01277599980433782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,65536,0.5125466452704536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,16384,0.2518746587965223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,128,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,51200,0.7538631227281359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,16384,0.13019466400146484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,12288,0.20740444130367705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,51200,0.38938222991095645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,10240,0.17294932736290825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,16384,0.110207113954756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,65536,0.4323155615064833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,10240,0.1182897753185696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,12288,0.1154515610800849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,51200,0.37320354249742294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,8192,0.13911822107103136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,10240,0.06819378005133735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,12288,0.08539555470148723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,7168,0.12238400512271458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,8192,0.05807199743058947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,6144,0.10494310988320245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,8192,0.08897511164347331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,7168,0.050092445479498975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,7168,0.07457688781950209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,65536,0.9687573115030924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,6144,0.044898665613598294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,4096,0.07214311096403334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,6144,0.0792177783118354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,5120,0.05615644322501289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,4096,0.03257511059443156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,3584,0.028959999481836956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,3584,0.04584622383117676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,3072,0.05508444375462002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,2560,0.04604355494181315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,4096,0.05123200019200643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,3072,0.026156443688604567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,2048,0.03692533241377936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,3072,0.03847200009557936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,2560,0.022700443863868713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,1536,0.02860800094074673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,2560,0.030569778548346624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,2048,0.019458666443824768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,1024,0.02032088902261522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,2048,0.02906844351026747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,1536,0.016377778516875375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,5120,0.08867110808690389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,768,0.015135111080275642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,1536,0.02292622294690874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,1024,0.013420444395807056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,512,0.010676444404655032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,1024,0.016807110773192514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,3584,0.0629262195693122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,256,0.0064195555945237475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,768,0.01349422252840466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,512,0.011038222246699862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,768,0.014125333891974555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,512,0.010999999940395355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,256,0.009399111072222391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,64,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,32,0.0028088889602157804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,256,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,128,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,128,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,64,0.008403555386596257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,32,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,65536,0.32695645756191677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,65536,0.173905783229404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,51200,0.2596195538838704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,65536,0.15422933631473118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,16384,0.088755554623074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,51200,0.13088266054789224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,12288,0.06910577747556898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,16384,0.04182666540145874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,10240,0.05150755577617221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,16384,0.04786755641301473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,51200,0.12627289030287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,12288,0.03333244389957852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,12288,0.04805066519313388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,8192,0.04375466704368591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,7168,0.038104888465669416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,8192,0.025637333591779072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,10240,0.04429777794414096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,6144,0.03219199842876858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,7168,0.022005332840813532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,7168,0.030794666873084173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,8192,0.035409778356552124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,5120,0.028000000450346205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,4096,0.023563555545277063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,6144,0.019552888141738046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,5120,0.017775111728244357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,6144,0.029081778393851385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,3584,0.020232000284724765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,5120,0.023452444208992854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,4096,0.015331554743978711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,3072,0.017775999175177682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,4096,0.02167111138502757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,2560,0.015124445160230001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,3584,0.014359111587206522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,3584,0.019148444135983784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,3072,0.01348355578051673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,2048,0.012338666452301873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,2560,0.012161778079138862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,1536,0.010011555420027839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,2048,0.010836444795131683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,5120,0.03761777612898085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,1024,0.007249777515729268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,128,0.0031946665710873077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,1536,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,768,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,1536,0.01144977741771274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,1024,0.007824000385072496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,10240,0.029542220963372126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,512,0.004197333421972063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,1024,0.009329777624871995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,768,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,256,0.0025031111306614345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,768,0.007502222226725684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,512,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,128,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,512,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,256,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,64,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,256,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,32,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,128,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,64,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,3072,0.017024889588356018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,65536,0.25193243556552464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,2560,0.013983110586802164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,2048,0.013608000344700284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,51200,0.19807555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,65536,0.15181777212354872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,65536,0.13152889410654703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,16384,0.06558222240871854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,51200,0.10690844721264309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,12288,0.05038933290375603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,16384,0.037485334608289934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,10240,0.042845331960254245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,12288,0.03425777620739407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,128,0.006747555401590135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,12288,0.04103199972046746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,8192,0.03590222199757894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,10240,0.025780444343884785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,10240,0.03762311074468825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,7168,0.03129333257675171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,32,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,8192,0.022237333986494277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,8192,0.030984888474146526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,6144,0.027303111222055223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,5120,0.023794665932655334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,7168,0.019861333900027804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,7168,0.026558222042189703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,6144,0.017883555756674874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,4096,0.01959911154376136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,6144,0.02433066732353634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,5120,0.016192888220151264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,3584,0.016545777519543965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,5120,0.020228443874253165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,4096,0.014140443669425117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,3072,0.014558222558763293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,4096,0.01965777741538154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,2560,0.012319111161761813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,16384,0.041208889749315046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,3584,0.013464888764752282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,3072,0.012453333371215396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,2048,0.010689777632554373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,3072,0.015459555718633862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,2560,0.011001778145631155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,1536,0.008444444172912175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,2048,0.011580444044537015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,2560,0.013040000365840064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,1024,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,2048,0.012413333687517377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,1536,0.008541333178679148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,768,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,1536,0.010682666467295753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,512,0.0038355555799272326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,768,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,768,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,256,0.002887111157178879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,512,0.0070568886068132185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,512,0.006218666831652324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,128,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,256,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,64,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,256,0.005633777628342311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,51200,0.11225956016116673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,32,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,128,0.00654755574133661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,128,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,3584,0.017125333348910015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,64,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,65536,0.2159271107779609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,32,0.005695111221737332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,16384,0.05784533421198527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,51200,0.1710613303714328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,65536,0.13902933067745632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,65536,0.12709777885013157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,51200,0.10730222198698257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,1024,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,16384,0.03548089000913832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,51200,0.09736800193786621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,16384,0.03930133250024583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,1024,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,8192,0.030768887864218816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,12288,0.03651644455062018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,12288,0.030545777744717065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,7168,0.027484445108307734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,10240,0.02461066676510705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,6144,0.024365334047211543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,10240,0.031038221385743883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,8192,0.026696888936890498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,7168,0.018858666221300762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,5120,0.020854221449957952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,7168,0.02277155551645491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,6144,0.017186666528383892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,4096,0.01753777762254079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,6144,0.020828444096777175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,5120,0.015416888727082146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,5120,0.017723555366198223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,3584,0.015040889382362366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,4096,0.013802666631009845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,3072,0.013470222552617392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,3584,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,3584,0.014766222900814481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,12288,0.04478133387035794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,2560,0.011404444773991903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,3072,0.011671110987663269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,3072,0.013351110948456658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,2048,0.009906666974226633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,2560,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,10240,0.03667644328541226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,1536,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,2560,0.011545778148704104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,2048,0.009011555876996782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,1024,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,1536,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,2048,0.01129955550034841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,768,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,8192,0.02086755633354187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,1024,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,512,0.003842666745185852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,1024,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,768,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,256,0.0028728888266616394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,512,0.006653333289755716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,128,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,512,0.006292444550328785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,256,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,64,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,256,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,32,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,4096,0.016116445263226826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,128,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,64,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,65536,0.17180089155832926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,32,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,51200,0.1285626623365614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,65536,0.09310133589638604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,65536,0.07253422339757283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,16384,0.04392533169852363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,51200,0.06871733400556776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,12288,0.03453155689769321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,51200,0.05992444356282552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,1536,0.009300444689061906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,10240,0.02663733396265242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,16384,0.027857777145173814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,768,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,12288,0.0199635558658176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,12288,0.02813777658674452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,10240,0.016682666209008958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,7168,0.02077155477470822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,10240,0.024099555280473497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,8192,0.014658666319317289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,6144,0.018030222919252183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,8192,0.02035288843843672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,5120,0.01551644504070282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,7168,0.01407199932469262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,7168,0.018046221799320646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,6144,0.012701333396964602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,5120,0.011920889218648275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,6144,0.017102221647898357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,3584,0.01166044423977534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,5120,0.014264000786675347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,4096,0.010975110861990186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,4096,0.013275555438465543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,3584,0.009684444301658207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,3584,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,16384,0.022638221581776936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,3072,0.00869066682126787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,3072,0.011346666349305047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,2048,0.007585778004593319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,2560,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,8192,0.02400622268517812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,1536,0.006342222293217977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,2560,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,2048,0.0073724447025193115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,128,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,2048,0.00924177798959944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,1536,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,768,0.0038542221817705366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,1536,0.007899555895063613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,1024,0.006561777657932705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,512,0.0028693332440323303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,4096,0.014111111561457316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,1024,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,768,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,256,0.002493333278430833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,768,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,128,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,512,0.005986666513813867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,512,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,3072,0.010527110762066312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,256,0.005306666509972678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,128,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,32,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,128,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,2560,0.008976000050703684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,65536,0.1469795571433173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,64,0.005231110999981563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,32,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,51200,0.11487821737925212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,65536,0.06583822435802884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,16384,0.040074666341145836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,1024,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,51200,0.053429332044389516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,51200,0.06279200315475464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,12288,0.031622221072514854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,10240,0.0266942216290368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,16384,0.02207999924818675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,16384,0.02604177759753333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,12288,0.017895110779338412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,8192,0.022963555322753057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,10240,0.015577778220176697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,7168,0.019561777512232464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,8192,0.013854222165213691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,10240,0.02261688808600108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,6144,0.016872889465755887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,256,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,8192,0.019878221882714164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,5120,0.014760888285107084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,6144,0.01204711116022534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,6144,0.016511110795868766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,4096,0.012791111237472959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,5120,0.011036444041464063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,5120,0.013887999786270989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,3584,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,4096,0.009527111219035255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,65536,0.08269511328803168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,4096,0.012479110724396177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,3072,0.009780444204807281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,3584,0.008796444369686974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,2560,0.008579555484983657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,3584,0.011372444530328115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,2048,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,2560,0.007576888634098901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,3072,0.010751999914646149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,1536,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,2560,0.009453333086437648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,12288,0.025965332984924316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,1024,0.004482666651407878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,2048,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,1536,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,768,0.003519999898142285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,1536,0.007335999773608313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,7168,0.012832889126406776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,512,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,1024,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,7168,0.017112000121010672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,768,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,768,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,512,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,128,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,512,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,256,0.005655999812814925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,256,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,64,0.0018782222436534034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,32,0.0018773333479960759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,128,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,128,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,65536,0.12120089266035293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,3072,0.008006222546100616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,64,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,32,0.0053395554423332214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,51200,0.0957697762383355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,16384,0.03393066591686673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,65536,0.07621955871582031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,65536,0.06344266732533772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,2048,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,12288,0.02689244515366024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,51200,0.056516442033979625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,51200,0.05144977900716993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,10240,0.023077333966890972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,16384,0.0211173329088423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,1024,0.006231110956933763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,16384,0.023715555667877197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,12288,0.01715022159947289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,7168,0.017606221967273288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,10240,0.015212444795502557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,256,0.0022897778285874263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,10240,0.02095555596881443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,6144,0.015484443969196744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,8192,0.018844443890783522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,7168,0.012712000144852532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,5120,0.013756444056828817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,6144,0.011735999749766456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,4096,0.010718222293588849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,6144,0.014968888627158271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,5120,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,3584,0.009311999711725447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,5120,0.01331555595000585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,4096,0.008996444443861643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,3072,0.009108444054921469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,4096,0.01219377749496036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,3584,0.008015111088752747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,3584,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,2560,0.00796622203456031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,3072,0.007823111282454597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,3072,0.010269333091047075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,2048,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,2560,0.007338666253619724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,8192,0.019896888070636325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,2560,0.008612444831265343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,1536,0.005863111052248213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,2048,0.0070062221752272705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,12288,0.023352889551056757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,2048,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,8192,0.013462222284740873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,1536,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,768,0.00327822214199437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,1536,0.00721777809990777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,7168,0.01773866679933336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,768,0.006025777922736273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,1024,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,256,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,768,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,512,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,128,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,512,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,256,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,64,0.0019306666735145782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,256,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,128,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,128,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,64,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,65536,0.10664800140592788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,32,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,51200,0.0840266678068373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,1024,0.004148444367779626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,65536,0.061259554492102734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,65536,0.06697866651746961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,16384,0.030468444029490154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,12288,0.02436711059676276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,1024,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,51200,0.051186667548285596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,51200,0.049518220954471164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,16384,0.021348444951905146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,12288,0.01666844387849172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,8192,0.018147556318177115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,12288,0.022630222969584998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,10240,0.014852444330851236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,10240,0.020294222566816542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,7168,0.015795555379655626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,8192,0.013292444248994192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,32,0.0019137778629859288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,6144,0.014127999544143677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,8192,0.017161778277821012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,7168,0.012404444317022959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,5120,0.012238221863905588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,7168,0.01550044450494978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,6144,0.011370666325092316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,4096,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,5120,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,6144,0.014318222800890604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,3584,0.009424000150627559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,4096,0.008661333057615492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,3072,0.008603555460770925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,4096,0.011915555430783166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,3584,0.010665777656767102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,2560,0.007469333708286285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,3072,0.007772444850868649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,512,0.0028275555620590844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,3072,0.009222221871217092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,10240,0.02070755594306522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,2560,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,1536,0.005304000029961268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,2560,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,2048,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,1024,0.0038542221817705366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,2048,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,1536,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,768,0.0034888887570963967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,1536,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,512,0.0028408887899584244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,1024,0.006072000082996156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,768,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,256,0.00240622212489446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,768,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,512,0.0053688887920644544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,128,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,5120,0.012774222426944308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,256,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,64,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,3584,0.007962666451931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,256,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,32,0.0018684443914228016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,128,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,65536,0.08517955409155951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,128,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,64,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,16384,0.01997422178586324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,32,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,65536,0.06126755475997925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,65536,0.05874400006400215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,16384,0.025003555748197768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,51200,0.04757066567738851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,12288,0.020089776979552377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,16384,0.019595555133289762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,16384,0.019215111931165058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,10240,0.017616889543003507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,12288,0.016179554992251925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,12288,0.020357333951526217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,512,0.005787555542257097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,10240,0.014291556345091926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,10240,0.01847999956872728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,7168,0.013012444807423485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,8192,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,8192,0.018940443793932598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,6144,0.011360000405046674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,7168,0.011398221883508893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,7168,0.014096889230940076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,51200,0.06593777736028035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,5120,0.00998755544424057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,6144,0.010201777848932479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,6144,0.013756444056828817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,4096,0.008990222381220924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,5120,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,3584,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,5120,0.01185155577129788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,51200,0.04652977652019925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,4096,0.008029333419269985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,3072,0.00702044450574451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,4096,0.011720889144473605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,2560,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,3072,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,3584,0.009309333231714036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,2048,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,3072,0.007682667010360294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,8192,0.015418666932317944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,1536,0.004028444488843282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,2560,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,2560,0.007675555845101674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,1024,0.003170666595300039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,1536,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,2048,0.007051555646790399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,768,0.003010666618744532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,1536,0.00665155549844106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,1024,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,1024,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,2048,0.006567111031876669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,768,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,256,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,512,0.005772444523043103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,512,0.0053004444473319584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,128,0.001957333336273829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,256,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,256,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,64,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,3584,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,128,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,32,0.0019386667344305252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,128,0.004860444615284602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,65536,0.07642310857772827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,64,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,51200,0.058600889311896436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,32,0.005136000199450387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,2048,0.006821333534187741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,16384,0.022064000368118286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,65536,0.05733244286643135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,65536,0.05801600217819214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,12288,0.017842666970358956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,51200,0.04680088824696011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,51200,0.04487288991610209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,10240,0.015332443846596612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,16384,0.018565333551830716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,12288,0.01585422290696038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,8192,0.01332800007528729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,12288,0.01965688831276364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,10240,0.014167111780908374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,7168,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,10240,0.017847999930381775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,8192,0.012127999630239276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,6144,0.011003555523024665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,8192,0.017159110969967313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,5120,0.009669333696365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,7168,0.013462222284740873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,6144,0.009992000129487779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,4096,0.00868711123863856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,6144,0.012431999875439538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,5120,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,5120,0.011048888994587792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,3584,0.0075902218619982404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,4096,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,3072,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,4096,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,3584,0.008042666647169325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,2560,0.006009777800904379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,3584,0.008294222255547842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,3072,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,2048,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,3072,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,16384,0.01924088928434584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,512,0.00254844439526399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,2560,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,1536,0.003866666720973121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,768,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,1024,0.0032044444233179092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,2048,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,1536,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,768,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,1536,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,1024,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,7168,0.011580444044537015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,512,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,1024,0.00566133318675889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,256,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,768,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,512,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,768,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,128,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,256,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,512,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,64,0.0018959999498393799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,32,0.0018844445132546956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,256,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,128,0.00574222207069397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,65536,0.06115644507937961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,128,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,64,0.004991111242108875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,51200,0.04897333184878031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,32,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,16384,0.01924622224436866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,65536,0.05714222457673815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,65536,0.05486933390299479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,2560,0.007238222493065729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,12288,0.015466666883892484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,51200,0.045829332537121244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,16384,0.018906666172875296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,51200,0.04303644431961907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,10240,0.015154666370815702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,12288,0.01552177800072564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,16384,0.017655111021465726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,12288,0.018248889181349013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,7168,0.012060444388124677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,8192,0.011898666620254517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,10240,0.015432889262835184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,6144,0.010860444770918952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,8192,0.013354666531085968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,7168,0.01036533299419615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,5120,0.009022222624884712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,6144,0.009675555759006077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,6144,0.012588444683286877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,4096,0.008330666356616551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,5120,0.00887822194231881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,3584,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,5120,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,4096,0.008272888759771982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,3072,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,4096,0.008368000388145447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,3584,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,2560,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,3584,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,3072,0.007682667010360294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,2048,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,2048,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,3072,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,1536,0.0038328886859946777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,2560,0.007096889118353526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,8192,0.013094222380055321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,2560,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,1024,0.0031893334041039148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,10240,0.013918222652541267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,2048,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,768,0.0028311111446883944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,1024,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,1536,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,512,0.002534222271707323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,1024,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,768,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,256,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,7168,0.013987556099891663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,512,0.005761777775155173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,128,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,512,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,64,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,128,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,256,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,32,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,128,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,65536,0.056919111145867236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,64,0.004987555659479565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,51200,0.04316088888380262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,65536,0.05590133534537422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,65536,0.052482667896482676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,16384,0.01959022217326694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,51200,0.04520533482233683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,2048,0.006654222392373615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,51200,0.04155288802252875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,12288,0.016224000189039443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,1536,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,10240,0.013926222920417786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,16384,0.01904177831278907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,16384,0.017160000072585214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,12288,0.015044444137149386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,8192,0.01274577776590983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,12288,0.01629244453377194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,10240,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,10240,0.014797333214018079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,7168,0.011170667078759937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,256,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,8192,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,6144,0.010115555591053432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,8192,0.013296888934241401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,5120,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,7168,0.01033688916100396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,7168,0.01163644426398807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,4096,0.00795911086930169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,6144,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,5120,0.008962666822804345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,3584,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,32,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,5120,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,3072,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,4096,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,4096,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,3584,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,2560,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,3072,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,3584,0.007248889240953658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,2048,0.004278222305907143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,3072,0.0069848886794514125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,2560,0.00721155520942476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,1536,0.003859555555714501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,2560,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,2048,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,1024,0.0031439999325407874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,1536,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,2048,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,768,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,1536,0.00609777785009808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,768,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,512,0.002499555548032125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,768,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,1024,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,256,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,512,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,768,0.005384888913896348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,128,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,512,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,256,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,6144,0.009376889301670922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,64,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,32,0.0018719999740521114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,256,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,128,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,65536,0.04811822374661764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,51200,0.035120887888802424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,64,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,32,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,65536,0.05509688787990146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,16384,0.016107555892732408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,65536,0.05065777897834778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,12288,0.013087111214796701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,51200,0.04467466804716322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,16384,0.01854311095343696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,10240,0.011830222275522021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,16384,0.016487111647923786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,12288,0.013790222505728403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,8192,0.011039999624093374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,1024,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,12288,0.015710221396552194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,10240,0.014296889305114746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,7168,0.009640889035330879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,8192,0.01238933371173011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,8192,0.01074488874938753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,7168,0.010085333552625444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,5120,0.008079110748238033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,7168,0.009732444253232744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,6144,0.009375111096435124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,4096,0.007273777491516537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,6144,0.00906755526860555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,128,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,5120,0.008781332936551835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,3584,0.006566222343179915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,4096,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,5120,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,3072,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,4096,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,2560,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,51200,0.04022399915589227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,3072,0.007344889144102733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,2048,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,3072,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,2560,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,1536,0.0038088887102074097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,2560,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,2048,0.006728888799746831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,1024,0.003126222226354811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,10240,0.012159110771285163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,2048,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,768,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,1536,0.006735111276308696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,1536,0.006091555373536215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,6144,0.008969777988062965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,512,0.0024666666156715816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,1024,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,768,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,768,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,256,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,512,0.005655111124118169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,128,0.001992000059949027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,512,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,256,0.005653333332803514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,64,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,256,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,32,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,3584,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,128,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,65536,0.03459999958674113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,64,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,51200,0.02741688821050856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,32,0.00508355556262864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,65536,0.05470400055249532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,16384,0.012448888685968188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,51200,0.04417511158519321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,51200,0.03872266742918227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,12288,0.011347555451922946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,16384,0.016701334052615695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,1024,0.005980444451173146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,10240,0.00981955561372969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,16384,0.016223111086421542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,12288,0.013470222552617392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,8192,0.008955555657545725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,12288,0.014186667071448432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,10240,0.012095111111799875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,7168,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,10240,0.01239733315176434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,8192,0.010479110810491772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,6144,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,128,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,8192,0.010073777702119615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,7168,0.010013333625263637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,5120,0.007124444676770105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,7168,0.009313777916961247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,3584,0.007143110864692264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,4096,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,6144,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,3584,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,6144,0.008856000171767341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,5120,0.008112000094519721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,4096,0.008285333712895712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,3072,0.005385777602593104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,4096,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,2560,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,3584,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,3072,0.00739911115831799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,2048,0.0041884444653987885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,3072,0.00699733363257514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,2560,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,1536,0.0035457776652442086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,2560,0.006639110959238476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,2048,0.0069013333155049225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,1024,0.003175111073586676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,2048,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,1536,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,768,0.0028702221396896574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,1536,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,1024,0.005966222120655908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,1024,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,768,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,256,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,768,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,512,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,512,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,128,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,5120,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,256,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,64,0.0018568889548381169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,256,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,128,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,65536,0.027759111589855615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,3584,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,128,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,51200,0.021918222308158875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,64,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,32,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,65536,0.054198222027884595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,65536,0.046690666013293795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,51200,0.04372266597217984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,12288,0.008522666990756989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,51200,0.03769866625467936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,16384,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,16384,0.01549600064754486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,10240,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,12288,0.013109332985348172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,512,0.0024968888610601425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,8192,0.007537778053018782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,12288,0.01273333364062839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,7168,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,10240,0.01202666676706738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,10240,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,8192,0.010695999695195092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,6144,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,7168,0.010033778018421596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,7168,0.009311999711725447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,32,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,5120,0.006365333166387346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,6144,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,4096,0.00656444455186526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,6144,0.008683555656009251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,65536,0.04864977796872457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,3584,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,4096,0.008031110796663497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,5120,0.00812977800766627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,16384,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,4096,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,3584,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,2560,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,3584,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,3072,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,2048,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,2560,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,2560,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,1536,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,2048,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,2048,0.006219555520349079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,1024,0.003640000190999773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,1536,0.0063866666621632045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,8192,0.00998399986161126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,1536,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,1024,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,1024,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,512,0.0026862221873468826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,768,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,5120,0.008612444831265343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,256,0.0021946666141351065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,512,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,3072,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,512,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,128,0.0021182222084866627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,256,0.005542222410440445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,64,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,256,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,128,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,32,0.0018604445374674266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,3072,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,65536,0.021919111410776775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,128,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,51200,0.01751733322938283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,64,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,32,0.004855111241340637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,65536,0.0539644459883372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,16384,0.00923733330435223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,65536,0.04616533385382759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,51200,0.04313866628540886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,12288,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,51200,0.037167112032572426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,768,0.0031893334041039148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,10240,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,16384,0.014718222949239941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,12288,0.01311377767059538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,8192,0.006224888894293044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,10240,0.01202133297920227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,12288,0.012416000167528788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,7168,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,768,0.0053022222386466126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,10240,0.011066666907734342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,8192,0.010393777655230628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,6144,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,5120,0.006335111127959357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,8192,0.009720000127951304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,7168,0.010084444450007545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,4096,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,6144,0.009300444689061906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,7168,0.009338666995366415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,5120,0.008661333057615492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,6144,0.00870400004916721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,4096,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,3072,0.0053671110007498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,5120,0.008011555506123437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,2560,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,4096,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,3584,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,3072,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,2048,0.004182222402758068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,2560,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,3072,0.00702311098575592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,1536,0.004527111020353106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,2560,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,2048,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,1024,0.0036186666952239144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,2048,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,1536,0.006593777901596493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,768,0.0033173333439562055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,16384,0.015855999456511605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,1536,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,512,0.0025537777692079544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,1024,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,768,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,256,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,512,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,768,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,128,0.0021297778520319196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,512,0.005095110999213325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,256,0.005514666852023866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,256,0.004965333475006951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,32,0.0018782222436534034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,128,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,65536,0.015628443823920358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,768,3584,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,128,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,51200,0.013082666529549493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,64,0.004853333450025982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,3584,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,32,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,65536,0.05077422327465481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,16384,0.007600000335110559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,12288,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,65536,0.04532266656557719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,51200,0.03847288754251268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,10240,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,51200,0.036624001132117376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,12288,0.013106666505336761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,16384,0.014514666464593677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,8192,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,12288,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,10240,0.011717333561844297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,7168,0.005336888962321811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,10240,0.011015999648306104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,8192,0.010429333481523724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,7168,0.009881777895821465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,8192,0.009763555394278632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,5120,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,7168,0.009331555830107795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,6144,0.009200000100665623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,4096,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,5120,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,5120,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,3584,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,4096,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,4096,0.007538666327794393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,3072,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,3584,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,3584,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,2560,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,3072,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,16384,0.015824889143308003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,3072,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,2048,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,2560,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,1536,0.0036640001667870414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,2560,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,2048,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,1024,0.006051555689838197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,1024,0.002883555574549569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,6144,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,2048,0.005976888868543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,768,0.002834666727317704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,1536,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,512,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,1024,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,1536,0.005679111099905438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,256,0.0021991110924217436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,1024,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,6144,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,512,0.005535111245181825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,768,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,128,0.0020346666375796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,64,0.0018595556418100994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,512,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,256,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,512,32,0.0018213333355055915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,65536,0.010732444624106089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,256,0.004860444615284602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,128,0.004566222015354368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,64,0.004605333424276776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,65536,0.0461297763718499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,32,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,16384,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,65536,0.042816887299219765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,12288,0.005381333331267039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,51200,0.037068443165885076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,16384,0.015807110402319167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,51200,0.03429244293106927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,10240,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,16384,0.014505777094099255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,12288,0.013088000317414602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,8192,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,12288,0.012268444730175866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,7168,0.0053182223604785065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,8192,0.010672888822025724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,8192,0.009782222410043081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,768,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,6144,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,7168,0.010013333625263637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,5120,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,6144,0.009310222334331935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,128,0.005382222019963794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,4096,0.004968889057636261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,5120,0.008632889224423302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,5120,0.007939555578761632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,51200,0.009333333207501305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,3584,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,4096,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,4096,0.007376888559924231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,3072,0.004559999952713649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,3584,0.007661333514584436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,3584,0.007098666495747036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,2560,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,3072,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,10240,0.011739555332395764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,3072,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,2048,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,2560,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,10240,0.0107360002067354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,1536,0.0035084444615576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,2560,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,2048,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,1024,0.0028764444092909494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,1536,0.0064311110311084324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,7168,0.00959911114639706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,768,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,1536,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,1024,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,512,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,1024,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,6144,0.008361777497662438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,256,0.0021484444538752237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,512,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,768,0.005306666509972678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,128,0.0018906665758954154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,512,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,256,0.005541333307822545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,64,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,256,32,0.001806222316291597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,256,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,128,0.005294222384691238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,65536,0.008041777544551426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,128,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,51200,0.006735999965005451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,64,0.004863111095296012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,65536,0.04587199952867296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,32,0.004565333326657613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,16384,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,65536,0.04167644513977898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,12288,0.004973333328962326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,51200,0.03707999984423319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,51200,0.03401155604256524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,16384,0.015560888581805758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,10240,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,16384,0.014141332772043018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,8192,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,12288,0.012039999994966718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,2048,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,10240,0.01204177737236023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,7168,0.004964444372389051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,10240,0.010826666322019366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,6144,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,8192,0.01036533299419615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,768,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,8192,0.009655999640623728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,5120,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,7168,0.009354666703277165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,6144,0.0092284447617001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,4096,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,6144,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,5120,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,3584,0.004538666870858935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,4096,0.008006222546100616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,3072,0.004242666479614046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,5120,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,2560,0.004197333421972063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,4096,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,3584,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,2048,0.004639111045334074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,3072,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,3584,0.007142222589916653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,2560,0.006996444529957241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,3072,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,2048,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,2560,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,1024,0.002842666581273079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,2048,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,1536,0.006509333435032103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,768,0.002542222125662698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,12288,0.013064000341627332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,512,0.002489777695801523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,768,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,768,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,256,0.002157333410448498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,512,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,512,0.004941333499219683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,7168,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,128,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,256,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,64,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,256,0.005094222310516569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,128,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,32,0.001826666709449556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,65536,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,128,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,64,0.004956444518433677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,51200,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,32,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,16384,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,128,1536,0.0035013332962989807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,65536,0.042103111743927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,12288,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,10240,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,51200,0.033952001068327166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,8192,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,1024,0.005877333382765452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,12288,0.01205244412024816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,7168,0.004576888763242298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,6144,0.0042168887125121225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,1024,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,5120,0.00452888881166776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,8192,0.009686222506894005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,7168,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,4096,0.004253333227501975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,3584,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,5120,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,3072,0.0041804446114434134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,4096,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,2560,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,3584,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,2048,0.004543111142185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,3072,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,1536,0.0032791110376516977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,2560,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,1024,0.002879111096262932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,1536,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,16384,0.014442667365074158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,10240,0.010934222075674268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,512,0.0024542222834295696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,768,0.005154666801293691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,256,0.0021297778520319196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,6144,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,128,0.0018311111877361934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,512,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,64,0.0017866666118303935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,256,0.004849777867396672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,32,0.001718222267097897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,128,0.004876444323195351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,65536,0.00758755538198683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,32,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,2048,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,65536,0.04172622164090475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,16384,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,1536,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,12288,0.004941333499219683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,1024,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,51200,0.03397511111365424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,10240,0.004953777624501122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,8192,0.0047493332790003884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,16384,0.014138667119873894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,7168,0.004568888909286923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,6144,0.004579555657174853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,12288,0.01178311142656538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,5120,0.004615111069546806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,10240,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,8192,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,7168,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,6144,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,3584,0.004265777766704559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,3072,0.003920000046491623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,5120,0.007692444655630324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,2560,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,64,0.004868444469239977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,2048,0.00423555572827657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,3584,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,3072,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,1536,0.003552888830502828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,1024,0.0028719999310043124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,51200,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,2560,0.0063075555695427795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,768,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,512,0.0022008888837363985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,2048,0.006230222268237009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,256,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,64,768,0.002469333302643564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,1536,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,128,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,1024,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,64,0.0015795555793576771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,32,0.0017875555074877208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,768,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,512,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,256,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,128,0.004625777817434735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,64,0.005088888936572605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,32,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,10,32,4096,0.0041999999019834726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,16384,0.33045866754319936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,12288,0.25298134485880536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,4096,0.007651555869314406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,16384,0.16279644436306423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,51200,0.4873039987352159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,16384,0.13550844457414415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,10240,0.18716888957553438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,51200,0.9892817603217231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,12288,0.14430310991075304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,8192,0.1531182262632582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,10240,0.1237635612487793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,12288,0.10084266795052423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,10240,0.08716622326109146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,7168,0.13205777274237737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,8192,0.11344710985819499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,6144,0.11390488677554661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,8192,0.07078488667805989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,5120,0.09582755300733779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,7168,0.061999109056260854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,7168,0.09520888990826076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,6144,0.05383733246061537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,4096,0.07909511195288764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,51200,0.4342871242099338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,6144,0.08535999721950954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,5120,0.046911110480626426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,3584,0.06787377595901489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,5120,0.07437422540452746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,4096,0.03862311111556159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,3072,0.05888711081610786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,4096,0.06420444117652045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,2560,0.049534221490224205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,3584,0.03487199876043532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,2048,0.04054311248991225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,3072,0.03091644578509861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,3072,0.04810755451520284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,3584,0.05716533131069607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,1536,0.03129422333505418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,2560,0.04086755712827047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,2048,0.022991110881169636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,2048,0.034435556994544134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,768,0.01773155563407474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,1536,0.019425777925385367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,1024,0.01480888823668162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,1536,0.026311111119058397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,512,0.012630222572220696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,1024,0.019464888506465487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,256,0.007181333170996771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,768,0.015105777316623263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,512,0.012753778033786349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,768,0.016374222106403775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,128,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,512,0.01238044434123569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,256,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,64,0.004237333519591225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,32,0.0041857775714662345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,128,0.01036977767944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,128,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,256,0.009882666998439366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,64,0.008951110972298516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,32,0.009867555565304225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,2560,0.027127999398443434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,1024,0.022416889667510986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,16384,0.25146932072109646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,51200,0.38991112179226345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,51200,0.7528222401936849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,65536,0.9715315500895182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,12288,0.2066497802734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,16384,0.13023200300004747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,16384,0.10620444350772434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,65536,0.44170665740966797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,12288,0.11731378237406413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,51200,0.3427920076582167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,10240,0.1731955607732137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,10240,0.11346488528781468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,8192,0.13958844873640272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,65536,0.5159777535332574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,7168,0.12290222114986843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,12288,0.0854497750600179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,10240,0.07174222336875068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,6144,0.10447733269797431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,8192,0.057583113511403404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,8192,0.08904888894822861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,5120,0.08842133151160346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,4096,0.07315911187065972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,7168,0.05014400018586052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,5120,0.0384479992919498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,7168,0.07642310857772827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,3584,0.06292177571190728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,5120,0.061794665124681264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,3072,0.054732445213529796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,4096,0.03267911076545715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,2560,0.045976890458001025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,3584,0.028786665863460962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,3584,0.04583644535806444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,3072,0.025498666697078284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,2048,0.03696622120009528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,3072,0.04106400079197354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,2560,0.02784444557295905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,1536,0.028805332051383123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,2048,0.019486222002241347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,2560,0.03435733252101474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,2048,0.02889066603448656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,1024,0.02075466679202186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,768,0.014952000644471912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,1536,0.022282666630215112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,1024,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,512,0.01069155583779017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,1024,0.016525333126386006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,768,0.01328000012371275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,768,0.014131555954615274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,256,0.006406222366624408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,512,0.01090666651725769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,6144,0.04576444294717577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,512,0.011052444577217102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,128,0.0033280000918441345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,256,0.009416888985368941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,4096,0.05041155550214979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,256,0.00832266691658232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,64,0.0026568888376156488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,32,0.002854222224818336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,128,0.007301333049933116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,64,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,32,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,65536,0.32562488979763454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,65536,0.17382578055063883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,1536,0.016379555066426594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,51200,0.2605991098615858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,16384,0.08898666832182144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,65536,0.15591555171542698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,51200,0.13066666656070286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,6144,0.06798666715621948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,51200,0.1213715606265598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,16384,0.04201155569818285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,10240,0.051669332716200084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,128,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,12288,0.0333591103553772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,8192,0.04346488912900289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,10240,0.02827022141880459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,10240,0.04070933328734504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,7168,0.036972443262736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,8192,0.024838222397698298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,8192,0.0352151095867157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,6144,0.032107555203967623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,7168,0.02163822286658817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,5120,0.027541332774692114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,7168,0.03187288840611776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,6144,0.020224889119466145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,6144,0.027268444498380024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,4096,0.023166222704781428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,5120,0.02421066661675771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,5120,0.017451556192504037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,3584,0.020092444287406076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,12288,0.06896799802780151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,4096,0.015976889265908133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,3072,0.017851556340853374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,4096,0.021479111578729417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,2560,0.015184000134468079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,3584,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,2048,0.012392000191741519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,3072,0.016176000237464905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,12288,0.04798310995101929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,3072,0.01754755609565311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,2560,0.012015110916561551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,1536,0.00998133338159985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,1024,0.00740799970097012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,2048,0.010767111347781287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,1536,0.009163555171754625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,2048,0.013439111411571503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,768,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,1024,0.007788444558779399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,1536,0.011522666447692446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,512,0.004190222256713443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,1024,0.009176000124878353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,768,0.007962666451931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,512,0.0070977773931291364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,768,0.007540444533030192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,128,0.0025902222841978073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,512,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,256,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,256,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,128,0.006694222076071634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,32,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,16384,0.047425776720047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,64,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,3584,0.019567110472255282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,32,0.006223111102978389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,65536,0.25226932101779515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,2560,0.015089778436554803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,65536,0.14673156208462185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,51200,0.19836800628238252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,16384,0.06526488727993436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,65536,0.13890577687157526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,256,0.002571555475393931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,51200,0.11230666107601589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,12288,0.05004177822007073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,51200,0.10382222466998631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,16384,0.04145688811937968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,12288,0.029401779174804688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,64,0.0021182222084866627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,12288,0.04183377822240194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,8192,0.03566311134232415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,10240,0.025966222087542217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,7168,0.031005332867304485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,128,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,10240,0.0344453321562873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,6144,0.02706844442420536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,8192,0.021852445271280076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,5120,0.0236488895283805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,8192,0.031839112440745033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,6144,0.018163555198245578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,6144,0.023959111836221483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,4096,0.01919911139541202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,5120,0.016220443778567843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,5120,0.02160177793767717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,3584,0.01665244499842326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,4096,0.014410666293568082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,4096,0.01941777765750885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,3072,0.014737778239780001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,3584,0.013253333667914072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,16384,0.03708622190687392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,10240,0.042818665504455566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,2560,0.012242666549152799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,3072,0.012406222522258759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,3072,0.015783111254374187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,2048,0.010398222340477837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,2560,0.011048888994587792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,1536,0.008365333080291748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,2048,0.009736888938479954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,2560,0.014442667365074158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,1024,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,2048,0.012700444294346703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,7168,0.019796444310082328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,1536,0.008646222452322641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,768,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,7168,0.026910222238964502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,1536,0.010405333505736457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,1024,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,512,0.0038613333470291565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,256,0.002834666727317704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,768,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,1024,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,128,0.0024782222592168385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,512,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,768,0.0069955554273393415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,64,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,512,0.006622222148709827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,32,0.002161777681774563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,256,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,128,0.006517333288987477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,128,0.005208000126812193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,3584,0.01773244473669264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,64,0.005716444303592046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,32,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,51200,0.1709902286529541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,65536,0.14066755771636963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,16384,0.057392888598971896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,65536,0.12830311722225612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,51200,0.10732355382707383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,51200,0.09999555349349976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,12288,0.04469155602984958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,16384,0.035924444595972695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,10240,0.03665511144532098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,16384,0.03922755519549052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,8192,0.030505776405334473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,10240,0.0244862222009235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,10240,0.030737777551015217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,7168,0.027318222655190363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,8192,0.02090222140153249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,8192,0.025599999560250178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,6144,0.024413333998786077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,7168,0.01910044418440925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,5120,0.02071199980047014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,7168,0.023553777072164748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,65536,0.21562577618492973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,6144,0.017269333203633625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,4096,0.017449777987268236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,5120,0.015397333436542086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,3584,0.015118221441904703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,5120,0.017809778451919556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,4096,0.0140337778462304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,4096,0.016561778055297006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,3072,0.013408888545301227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,3584,0.0140364451540841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,2560,0.01142133358452055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,3584,0.0148062225845125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,3072,0.012040889097584618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,12288,0.027892443868849013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,2048,0.009672000176376766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,3072,0.013737777868906656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,12288,0.03483555714289347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,1536,0.008046222229798635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,2560,0.01223111069864697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,2048,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,1024,0.006603555546866522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,256,0.006593777901596493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,2048,0.011205332974592844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,1536,0.008090666598743862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,1536,0.009420444567998251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,1024,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,512,0.0038942222793896994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,1024,0.00756533361143536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,768,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,256,0.0028862222615215513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,768,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,512,0.006664888726340399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,128,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,512,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,256,0.006360000206364526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,64,0.0021315556433465746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,256,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,128,0.0063271112740039825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,32,0.0021964444054497615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,128,0.005299555758635203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,64,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,65536,0.1716773377524482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,32,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,51200,0.12809333536359999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,2560,0.010634666515721215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,65536,0.09291111098395453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,65536,0.07108266486061944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,16384,0.04376177655325996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,51200,0.06895466645558675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,12288,0.03453955716556973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,768,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,16384,0.02295111119747162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,10240,0.026551110876931086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,51200,0.0563102232085334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,16384,0.0273964438173506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,12288,0.01825066738658481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,8192,0.023919110496838886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,12288,0.027444443768925134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,7168,0.02065955599149068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,10240,0.01647911138004727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,6144,0.01819022165404426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,8192,0.014464888307783337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,10240,0.02457333273357815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,5120,0.015475556254386902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,8192,0.02013777858681149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,7168,0.01832088828086853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,6144,0.012494222157531314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,4096,0.013086222112178802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,6144,0.016597333881590102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,3584,0.01146933353609509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,5120,0.014087999860445658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,4096,0.010537777509954242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,3072,0.010416888528399997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,4096,0.012920888761679331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,3584,0.010346666806273991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,2560,0.008999110923873054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,3584,0.012642666697502136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,3072,0.008871999879678091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,2048,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,2560,0.009015999734401703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,2560,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,1536,0.006537777682145436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,2048,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,2048,0.009189333352777693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,1024,0.004842666702138053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,1536,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,6144,0.019905777441130746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,1536,0.00739644467830658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,768,0.0035333335399627686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,7168,0.013727111121018728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,1024,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,512,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,1024,0.006303999986913469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,256,0.002471111093958219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,768,0.0063164445261160536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,5120,0.011755555868148804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,512,0.005992888990375731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,256,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,512,0.005807999935415056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,64,0.0019911111642917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,128,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,32,0.0021893332401911416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,128,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,65536,0.14762132697635227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,64,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,32,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,51200,0.11464533540937637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,65536,0.08327822552786933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,16384,0.04004089037577311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,65536,0.06600977977116902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,51200,0.06262666649288602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,51200,0.05358044306437174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,12288,0.03151911165979173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,16384,0.02148977749877506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,10240,0.026445334156354267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,16384,0.02551555633544922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,12288,0.017606221967273288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,12288,0.026308443811204698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,8192,0.02276533345381419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,768,0.006261333409282897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,7168,0.019479110836982727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,10240,0.015772443678643968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,256,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,8192,0.013946667313575745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,10240,0.02186222208870782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,6144,0.017064889272054035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,5120,0.014707555373509726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,7168,0.013047999805874295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,7168,0.017824000782436795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,3072,0.011388444238238864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,4096,0.01276177747382058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,5120,0.011071999867757162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,6144,0.015665777855449252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,3584,0.011018666956159802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,5120,0.013521778086821238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,4096,0.009650666680600908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,3072,0.0099973330895106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,3584,0.008974221845467886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,3584,0.011702222128709158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,2560,0.008609777523411645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,3072,0.008565333154466417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,3072,0.010988444089889526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,2048,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,2560,0.007655999726719326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,2560,0.010000000397364298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,128,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,1536,0.00620977787507905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,2048,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,2048,0.008586666650242275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,1024,0.004507555729813046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,1536,0.006647999915811751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,768,0.0035111109415690103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,8192,0.019889778561062284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,1536,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,1024,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,512,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,1024,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,6144,0.012015110916561551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,256,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,768,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,128,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,512,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,768,0.005978666659858491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,256,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,64,0.0019075555933846368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,512,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,32,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,128,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,4096,0.01239466667175293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,128,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,65536,0.12124355634053548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,64,0.005207111024194294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,32,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,51200,0.09544710980521308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,16384,0.03391644358634949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,65536,0.06309333112504747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,65536,0.07511199845208062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,12288,0.027070222629441157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,51200,0.05140977766778734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,51200,0.05646489063898722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,16384,0.020814221766259935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,10240,0.02308711078431871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,12288,0.01718577742576599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,8192,0.0200311111079322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,16384,0.02328799996111128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,7168,0.017528888252046373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,12288,0.02369511127471924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,10240,0.015364444918102689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,8192,0.013444444371594323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,6144,0.015466666883892484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,10240,0.020566221740510728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,5120,0.013766222529941134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,8192,0.018569777409235638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,7168,0.012424888710180918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,4096,0.01069777790043089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,7168,0.016505777835845947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,6144,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,3584,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,5120,0.01038933296998342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,6144,0.014539556370841132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,3072,0.009110222260157267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,4096,0.009264888862768808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,5120,0.013348444468445249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,256,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,2560,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,4096,0.012100444071822695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,2048,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,3072,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,3584,0.011392888923486074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,1536,0.0058782220714622075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,2560,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,3072,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,1024,0.003945777813593547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,2560,0.00904088881280687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,2048,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,768,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,2048,0.007696000238259633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,1024,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,1536,0.0069066666894488865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,512,0.0028435554769304064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,256,0.002471999989615546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,768,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,1024,0.006224888894293044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,128,0.0021262222694026097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,768,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,512,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,256,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,512,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,32,0.001869333287080129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,128,0.005746666755941179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,256,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,128,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,65536,0.10605244504080878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,64,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,32,0.005085333353943295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,65536,0.06115111377504137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,16384,0.03045511245727539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,51200,0.0510062211089664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,51200,0.04969155457284716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,12288,0.024502222736676533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,16384,0.02025422288311852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,16384,0.021368889345063105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,1536,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,10240,0.02060622142420875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,12288,0.016507556041081745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,12288,0.02635555631584591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,10240,0.015000000596046448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,10240,0.01906488835811615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,8192,0.013080000049538083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,8192,0.017895110779338412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,7168,0.01240799989965227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,51200,0.08376622200012207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,6144,0.014128888646761576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,7168,0.01590844492117564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,5120,0.012309333516491784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,65536,0.06708266337712605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,3584,0.008275555239783393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,4096,0.011064888702498542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,6144,0.013814222481515674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,5120,0.012515555653307172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,3584,0.00962222201956643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,4096,0.008616888688670265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,4096,0.011609777808189392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,3072,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,3584,0.007932444413503012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,8192,0.018106665876176622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,2560,0.007452444069915348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,3072,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,3072,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,2048,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,7168,0.01609155535697937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,2560,0.0087013335691558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,1536,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,2048,0.0069004446268081665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,2048,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,1024,0.0038808890514903595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,1536,0.006504000061088138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,1536,0.00702044450574451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,6144,0.011414222419261932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,768,0.003485333174467087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,512,0.002871111035346985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,5120,0.009887110855844285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,1024,0.006018666757477655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,768,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,256,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,512,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,768,0.005666666560702854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,128,0.0021333334346612296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,512,0.005428444594144821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,256,0.005693333430422678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,64,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,32,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,256,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,128,0.005746666755941179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,3584,0.011039999624093374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,128,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,65536,0.08491288953357273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,64,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,51200,0.06566666894488864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,16384,0.025026667449209426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,32,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,2560,0.007254222200976477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,65536,0.058966219425201416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,12288,0.020077332854270935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,51200,0.047445333666271634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,10240,0.017549332645204332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,12288,0.016132444143295288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,16384,0.019136889113320243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,8192,0.015506666567590503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,12288,0.024127110838890076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,10240,0.014382221632533602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,10240,0.01776711146036784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,7168,0.012802666260136498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,1024,0.006228444476922353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,8192,0.01276533305644989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,6144,0.011354666617181567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,8192,0.016175111134847004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,7168,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,5120,0.010024000373151567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,6144,0.010371555884679159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,7168,0.014455111490355598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,4096,0.008952000074916417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,5120,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,6144,0.013258666627936892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,3584,0.007679999702506595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,5120,0.01237511138121287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,3072,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,3584,0.007711999946170383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,3584,0.009767110976907942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,2560,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,3072,0.007347555624114142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,65536,0.06039200226465861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,2048,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,3072,0.008960888617568547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,51200,0.04670133193333944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,16384,0.019539554913838703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,2560,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,1536,0.004004444513056013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,2048,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,1024,0.0032017777363459268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,2048,0.0070471109615431885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,1536,0.00665155549844106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,768,0.0028159999185138275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,1536,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,1024,0.006024888820118374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,512,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,1024,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,768,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,256,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,768,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,512,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,128,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,512,0.005389333185222413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,256,0.005543111099137201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,64,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,256,0.005139555368158552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,32,0.0018862223045693503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,128,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,128,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,65536,0.0758097767829895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,64,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,32,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,51200,0.05875733163621691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,65536,0.058063109715779625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,65536,0.05791288614273071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,16384,0.02196088929971059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,51200,0.046626667181650795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,51200,0.044756445619795054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,12288,0.017543110582563613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,16384,0.019223110543357003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,10240,0.01516266663869222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,16384,0.018236445056067575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,12288,0.01595288846227858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,8192,0.013178666432698568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,12288,0.019294222195943195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,10240,0.01717511150572035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,7168,0.011940444509188334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,8192,0.012095111111799875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,8192,0.014430221584108142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,4096,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,6144,0.011003555523024665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,7168,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,5120,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,7168,0.013890667094124688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,6144,0.009993777506881291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,4096,0.00848444468445248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,6144,0.012702222499582501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,5120,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,3584,0.0075911109646161394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,2560,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,5120,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,3072,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,4096,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,4096,0.009311111436949836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,2560,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,3584,0.00795377790927887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,3072,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,2048,0.004980444494220945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,3584,0.008997333546479544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,1536,0.004184000194072723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,2560,0.007248000138335758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,2560,0.007675555845101674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,1024,0.003162666741344664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,1536,0.006551110910044775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,2048,0.0070159998204973005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,768,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,1536,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,4096,0.008283555507659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,1024,0.006223111102978389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,512,0.0024684444069862366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,1024,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,768,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,256,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,768,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,128,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,512,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,256,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,64,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,256,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,32,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,128,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,128,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,65536,0.06111911270353529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,64,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,51200,0.04885155624813504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,32,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,65536,0.054979556136661105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,3072,0.007873777714040544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,65536,0.05714044306013319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,16384,0.019104889697498746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,12288,0.015840000576443143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,2048,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,51200,0.045894222127066724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,10240,0.01459200017982059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,51200,0.04311644368701511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,10240,0.01511111193233066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,16384,0.018933332628673978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,8192,0.013404444687896304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,12288,0.015584889385435315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,10240,0.013792888985739814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,12288,0.01720533271630605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,7168,0.011991110940774282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,8192,0.011713777979214987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,10240,0.014789332946141561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,512,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,8192,0.014191110928853353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,7168,0.010247111320495605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,5120,0.008939555949634975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,7168,0.012096000214417776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,6144,0.009461333354314169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,4096,0.008193777667151557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,5120,0.009004444711738164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,6144,0.011132444772455426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,3584,0.007236444287829929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,5120,0.010405333505736457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,4096,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,3072,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,4096,0.008375110725561777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,3584,0.00795377790927887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,2560,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,3584,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,2048,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,3072,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,2560,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,1536,0.0038364442686239877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,2560,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,2048,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,1024,0.003195555466744635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,2048,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,16384,0.017539555827776592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,1536,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,768,0.0028702221396896574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,1536,0.006224888894293044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,512,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,1024,0.006136000156402588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,768,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,6144,0.010724444356229572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,256,0.0021946666141351065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,512,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,768,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,128,0.0018622221218215094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,512,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,256,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,64,0.0018604445374674266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,32,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,128,0.005672889037264719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,256,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,128,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,64,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,32,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,51200,0.04334311021698845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,3072,0.007380444142553542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,65536,0.05606310897403293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,16384,0.019482667247454327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,51200,0.04519111249181959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,16384,0.018633777896563213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,12288,0.016179554992251925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,51200,0.0413680010371738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,16384,0.017143110434214275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,12288,0.01513777838812934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,12288,0.01660977800687154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,1024,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,10240,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,10240,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,7168,0.011125333607196808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,8192,0.011101333631409539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,6144,0.010090666512648264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,8192,0.013024888932704926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,7168,0.010066666536860997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,65536,0.05614755551020304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,7168,0.01129955550034841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,6144,0.009431111315886179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,4096,0.007860444486141205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,6144,0.010390222072601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,5120,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,3584,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,65536,0.05267288949754503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,5120,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,4096,0.007985778152942657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,3072,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,4096,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,3584,0.00793422261873881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,2560,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,3072,0.007438222567240397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,10240,0.013795555465751223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,3584,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,2048,0.004569777597983678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,1536,0.003843555433882607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,3072,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,2560,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,1024,0.003216888962520493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,2048,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,2560,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,1536,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,768,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,2048,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,1024,0.006062222023804982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,512,0.002465777720014254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,1024,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,768,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,256,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,768,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,512,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,128,0.0018862223045693503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,5120,0.008955555657545725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,512,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,256,0.005388444496525659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,256,0.004847110973464118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,32,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,128,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,65536,0.04782488942146301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,64,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,32,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,51200,0.0351857774787479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,65536,0.05522844526502821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,65536,0.05069777700636122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,16384,0.016075554821226332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,8192,0.012757333616415659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,51200,0.04481955700450473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,1536,0.006076444354322221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,12288,0.01310755560795466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,51200,0.04012889001104567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,16384,0.018230222993426852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,10240,0.011792889071835412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,12288,0.014107555150985718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,8192,0.010792888700962067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,12288,0.014890667464998035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,10240,0.01240799989965227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,64,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,7168,0.00962933318482505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,10240,0.01385155568520228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,128,0.005747555444637935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,6144,0.008944888909657797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,8192,0.012078222301271228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,7168,0.010080888867378235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,5120,0.008107555409272512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,7168,0.01032266683048672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,4096,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,6144,0.008789333204428354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,5120,0.008969777988062965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,3584,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,5120,0.008328888979223039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,4096,0.008369777765538957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,3072,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,4096,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,3584,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,2560,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,3584,0.0074942219588491656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,3072,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,2048,0.004238222208287981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,3072,0.006993778049945831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,2560,0.007276444799370236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,1536,0.0038337777886125776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,16384,0.016715556383132935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,2560,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,1024,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,2048,0.00702311098575592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,2048,0.006247111078765657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,1536,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,768,0.0028168888141711554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,1024,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,8192,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,1024,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,768,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,256,0.0022195554855797025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,768,0.0052968888647026485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,6144,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,512,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,256,0.005628444254398346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,64,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,128,0.005543111099137201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,256,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,32,0.001855111163523462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,128,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,65536,0.03431466552946303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,64,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,32,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,51200,0.02718488872051239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,65536,0.0551093320051829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,16384,0.012413333687517377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,65536,0.04855288730727302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,51200,0.044180442889531456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,12288,0.011624000138706632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,51200,0.038675556580225624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,16384,0.01628977722591824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,10240,0.009750222166379293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,1536,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,12288,0.013528000149461957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,512,0.0025235555238193935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,8192,0.008784888519181145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,12288,0.014332445131407844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,7168,0.008330666356616551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,10240,0.01204533295498954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,128,0.0018746666610240936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,8192,0.010728889041476779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,6144,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,8192,0.00981422182586458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,7168,0.010048888623714447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,5120,0.007104000283612146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,7168,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,6144,0.00960355583164427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,4096,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,6144,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,5120,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,3584,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,5120,0.008276444342401292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,4096,0.008135110967689091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,3072,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,4096,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,3584,0.00795911086930169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,2560,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,3584,0.007303111255168915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,16384,0.016693333784739178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,3072,0.007697777615653143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,2048,0.0042133331298828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,3072,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,2560,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,1536,0.003548444559176763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,2048,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,2560,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,1024,0.002995555599530538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,2048,0.00610133343272739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,768,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,1024,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,10240,0.012126222252845764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,512,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,512,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,1024,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,256,0.0021439999755885866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,768,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,768,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,512,0.0058062221441004015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,128,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,256,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,512,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,64,0.0018773333479960759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,32,0.001790222194459703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,128,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,65536,0.027796443965699937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,128,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,64,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,32,0.004866666677925321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,51200,0.021697777840826247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,65536,0.054402665959464185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,65536,0.046894222497940063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,51200,0.04383377896414863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,51200,0.03759555684195624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,12288,0.008637333081828224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,1536,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,16384,0.016029333074887592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,16384,0.01480888823668162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,10240,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,1536,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,8192,0.007637333538797166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,12288,0.013222222526868185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,10240,0.011751111182901593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,12288,0.012734221915404001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,7168,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,10240,0.01162755572133594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,8192,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,6144,0.007473777565691207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,8192,0.009720889230569204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,5120,0.0063235556913746735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,256,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,6144,0.009379555781682333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,7168,0.009328000247478485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,4096,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,6144,0.008642666869693333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,3584,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,5120,0.008608000146018134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,5120,0.008075555165608725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,4096,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,4096,0.007446222007274628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,2560,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,16384,0.010211555494202508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,3072,0.007343110938866933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,2048,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,3072,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,2560,0.0069884442620807225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,1536,0.004599110947714912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,2560,0.006360000206364526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,2048,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,2048,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,1536,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,1536,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,768,0.003170666595300039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,1024,0.005895111295912001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,1024,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,512,0.0028577778074476454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,7168,0.00997688869635264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,768,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,256,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,768,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,512,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,512,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,3072,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,256,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,256,0.004866666677925321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,3584,0.007822222179836696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,64,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,32,0.0018604445374674266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,3584,0.007257777783605788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,65536,0.021479111578729417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,128,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,64,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,32,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,1024,0.00387999994887246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,65536,0.0539191100332472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,65536,0.04583644535806444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,51200,0.04279822111129761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,51200,0.037231110864215426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,12288,0.0079297779334916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,16384,0.014766222900814481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,16384,0.015826667348543804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,10240,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,12288,0.013106666505336761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,12288,0.012062222593360476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,8192,0.006130666782458623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,128,0.0021671110557185277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,10240,0.01202488856183158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,10240,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,7168,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,128,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,8192,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,6144,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,8192,0.009746666583749983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,7168,0.00999111102686988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,5120,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,7168,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,6144,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,4096,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,51200,0.017417778571446735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,5120,0.008624888956546783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,3584,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,5120,0.0081386665503184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,16384,0.009096889032257928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,3072,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,3584,0.007824000385072496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,4096,0.007376000285148621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,2560,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,3584,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,3072,0.007364444434642792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,2048,0.004173333446184794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,3072,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,1536,0.004659555438492033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,2560,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,2048,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,1024,0.00350577798154619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,1536,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,1536,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,768,0.00309955556359556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,1024,0.005985777825117111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,1024,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,512,0.0026586666289303037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,768,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,768,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,256,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,512,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,6144,0.0087013335691558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,128,0.001871111078394784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,256,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,4096,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,64,0.0017946666727463405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,768,32,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,256,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,128,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,65536,0.015584000282817416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,128,0.004761777818202972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,51200,0.01313333296113544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,64,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,2560,0.007187555233637492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,65536,0.05033333433998955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,2048,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,65536,0.04550666610399882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,51200,0.03834844297832913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,12288,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,51200,0.03643644518322415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,16384,0.015801777442296345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,10240,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,12288,0.013120888835854001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,8192,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,12288,0.012087110843923358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,10240,0.011738667057620155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,7168,0.005341333233647876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,10240,0.01089599976936976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,8192,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,6144,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,7168,0.0099973330895106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,7168,0.00903200027015474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,512,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,5120,0.0063235556913746735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,6144,0.00927022182279163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,4096,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,6144,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,5120,0.00849599970711602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,5120,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,3584,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,4096,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,32,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,4096,0.007080888582600488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,3072,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,3584,0.007675555845101674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,2560,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,3072,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,3584,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,16384,0.01445777714252472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,2048,0.004863111095296012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,2560,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,1536,0.003559111307064692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,2048,0.0068986668354935115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,2560,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,1024,0.0031973332580592898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,2048,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,8192,0.010011555420027839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,1536,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,768,0.0028026666906144884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,1536,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,512,0.002513777878549364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,1024,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,1024,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,768,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,256,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,768,0.005394666559166378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,512,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,128,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,256,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,512,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,64,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,256,0.005015111217896144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,512,16384,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,65536,0.010451555252075195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,128,0.004546666724814309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,51200,0.009321778184837764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,64,0.004527111020353106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,65536,0.04651555418968201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,32,0.004570666700601578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,16384,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,3072,0.006776888751321369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,65536,0.042817778057522245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,12288,0.00518488883972168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,51200,0.03719377848837111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,10240,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,51200,0.034279998805787824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,16384,0.015847111741701763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,8192,0.0053226666318045715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,12288,0.013062222136391534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,16384,0.014208889669842191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,7168,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,12288,0.012114666402339935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,8192,0.010432889064153036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,6144,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,8192,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,7168,0.010058666268984476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,5120,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,7168,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,6144,0.009325332939624786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,4096,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,6144,0.008638222184446123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,5120,0.008343111309740279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,3584,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,5120,0.008309333688682979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,3072,0.0046728890803125174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,4096,0.007918222082985772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,3584,0.007681777907742395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,4096,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,2560,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,3072,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,2048,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,3072,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,2560,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,1536,0.003531555748648114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,2560,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,2048,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,1024,0.0028435554769304064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,2048,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,1536,0.00630666646692488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,10240,0.011699555648697747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,768,0.0025351111673646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,1536,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,10240,0.010909332997269101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,512,0.002521777732504739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,1024,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,256,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,512,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,768,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,128,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,512,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,256,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,64,0.0018257778137922287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,256,0.00499288903342353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,128,0.005312888986534542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,256,32,0.0018257778137922287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,65536,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,128,0.004552000098758274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,3584,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,32,0.004852444347408083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,51200,0.007248889240953658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,65536,0.041554666227764554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,65536,0.04588977826966179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,16384,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,51200,0.037008888191647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,51200,0.033732444047927856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,12288,0.005187555733654234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,1024,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,16384,0.014476444986131458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,16384,0.01551644504070282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,10240,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,768,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,8192,0.0052915554907586836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,12288,0.013090666797426013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,12288,0.01201777739657296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,10240,0.011751111182901593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,7168,0.005177777674463061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,8192,0.010504888991514841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,6144,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,128,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,8192,0.009890666438473595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,7168,0.009993777506881291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,5120,0.004860444615284602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,7168,0.0090524446633127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,4096,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,5120,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,6144,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,3584,0.004863111095296012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,5120,0.008022222254011366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,4096,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,3072,0.004552888787455029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,3584,0.007686222592989604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,3584,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,2560,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,3072,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,3072,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,2048,0.004594666676388847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,2560,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,1536,0.0035342222286595237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,2560,0.006249777972698212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,2048,0.007000000112586551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,1024,0.0028799999919202593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,10240,0.01072622256146537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,1536,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,768,0.002522666628162066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,1536,0.005658666706747479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,1024,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,512,0.002469333302643564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,768,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,1024,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,6144,0.009368889033794403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,256,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,64,0.004573333180612988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,768,0.00519822206762102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,128,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,512,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,64,0.0017831110292010838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,512,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,128,0.0052622221410274506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,4096,0.007174222005738153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,256,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,65536,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,51200,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,64,0.004691555682155821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,16384,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,32,0.004579555657174853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,12288,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,65536,0.041745778587129384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,10240,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,51200,0.03394488824738397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,8192,0.004983110974232356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,16384,0.014239999983045789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,7168,0.004516444272465176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,12288,0.011765333513418833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,2048,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,6144,0.004583999928500918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,5120,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,8192,0.009693333672152625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,4096,0.003931555483076307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,7168,0.008953777452309927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,3584,0.004518222063779831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,5120,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,3072,0.0042408886882993905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,4096,0.007188444336255391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,2560,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,128,32,0.0018177777528762817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,3584,0.006899555524190267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,2048,0.004202666795916027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,128,0.004859555512666702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,3072,0.00648266656531228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,1024,0.00281866660548581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,2560,0.006478222293986215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,768,0.002505777817633417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,2048,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,512,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,1536,0.005677333308590784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,256,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,1024,0.005290666802061929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,768,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,128,0.0018746666610240936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,512,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,10240,0.010715555813577441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,64,0.0017671111143297618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,32,0.00179822225537565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,256,0.004600888739029567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,65536,0.008340444829728868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,128,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,6144,0.00871377769443724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,51200,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,64,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,16384,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,32,0.004559111264016894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,12288,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,256,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,65536,0.041469332244661115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,8192,0.0046693334976832075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,51200,0.03390489021937052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,16384,0.014130666851997375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,7168,0.004606222112973531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,12288,0.011755555868148804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,6144,0.004519999855094486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,5120,0.004254222330119875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,8192,0.009402666654851701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,4096,0.0038986665507157645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,7168,0.009126221968068017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,6144,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,3584,0.004218666503826777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,5120,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,4096,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,2560,0.004295999805132548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,3584,0.006744888921578725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,2048,0.003909333298603694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,3072,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,1536,0.00350577798154619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,2560,0.006138666636413998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,1024,0.002663111107216941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,2048,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,10240,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,768,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,1536,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,512,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,256,0.002120888895458645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,64,1536,0.0032382222513357797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,1024,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,128,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,64,0.0014720000326633453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,10240,0.010738666686746808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,512,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,32,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,256,0.00507733349998792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,128,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,64,0.004519111166397731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,32,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,16384,0.34083467059665257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,51200,0.4872808986239963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,16384,0.16205155849456787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,51200,1.018887095981174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,12288,0.24054845174153647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,768,0.005388444496525659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,16384,0.15826222631666395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,9,32,3072,0.004222222086456087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,12288,0.1032515565554301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,12288,0.1413902176751031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,51200,0.4586790932549371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,10240,0.14302310678693983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,8192,0.15333955817752415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,10240,0.08441333638297187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,8192,0.10691022210650974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,8192,0.06904711325963338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,12288,0.13036978244781494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,7168,0.13218755192226833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,16384,0.17402933703528511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,6144,0.11374221907721625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,10240,0.20232533084021675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,7168,0.09517600138982137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,7168,0.06295377678341336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,5120,0.09540355867809719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,6144,0.08264533016416761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,6144,0.053661333190070264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,7168,0.08268355660968356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,5120,0.0713813304901123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,4096,0.07964266671074761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,6144,0.07362755801942614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,5120,0.06363822354210748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,4096,0.06386844317118327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,3584,0.06776977909935845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,4096,0.03898400068283081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,10240,0.11296799447801377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,3584,0.055551111698150635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,4096,0.052763554784986705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,3584,0.03492710987726847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,8192,0.09429244200388591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,2560,0.04961511161592272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,3072,0.03096710973315769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,3072,0.048826667997572154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,3584,0.04856266578038534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,2048,0.04050044549836053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,5120,0.045864890019098915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,2560,0.04066666629579332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,2560,0.03363644414477878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,3072,0.042937778764300875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,1536,0.03145600027508206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,2048,0.03398133317629496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,2048,0.02455733385350969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,2560,0.03863999909824795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,2048,0.03409333361519708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,1024,0.022505778405401442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,1536,0.01921688848071628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,3072,0.05898221996095446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,1536,0.029370665550231934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,768,0.017704889178276062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,1024,0.019889778561062284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,1024,0.017623111605644226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,512,0.01254488858911726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,51200,0.5036933157179091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,768,0.015239111251301236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,1024,0.023136888941129048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,768,0.01646844380431705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,256,0.007237333390447829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,512,0.012836444709036084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,512,0.01274222218328052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,768,0.019509333703253005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,128,0.004963555683692296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,256,0.010065777434243096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,256,0.010355555348926121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,64,0.004197333421972063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,128,0.010429333481523724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,128,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,32,0.00424888895617591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,512,0.018060444129837882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,1536,0.027079999446868896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,128,0.010831111007266574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,64,0.009329777624871995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,32,0.010036444498433007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,65536,0.5129457579718696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,51200,0.38972976472642684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,51200,0.7481333414713541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,65536,0.9715297487046984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,256,0.014390221900410123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,16384,0.24958843655056426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,16384,0.1301057736078898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,65536,0.43083643913269043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,12288,0.19162044260236952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,12288,0.11450489362080891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,16384,0.13112444347805446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,12288,0.09108977847629124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,10240,0.16049777136908636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,10240,0.10178577899932861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,16384,0.14099021752675375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,10240,0.06852266523573133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,8192,0.1393937799665663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,12288,0.1075964437590705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,8192,0.08686400122112697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,7168,0.12262133757273357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,8192,0.05813066826926338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,10240,0.09106489022572835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,51200,0.32540800836351186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,6144,0.10531555281745063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,7168,0.05158222383923001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,7168,0.07509422302246094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,5120,0.08908000257280137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,6144,0.06890933381186591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,6144,0.04503200120396084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,7168,0.0682248870531718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,4096,0.07208800315856934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,5120,0.060581333107418485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,5120,0.03871555460823907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,4096,0.0324524442354838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,5120,0.05292710993025038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,4096,0.052490668164359204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,3584,0.062423110008239746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,4096,0.045421332120895386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,3072,0.05454755491680569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,3584,0.029211554262373183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,8192,0.07689688603083293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,3584,0.045307556788126625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,2560,0.04577066832118564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,3072,0.02551377813021342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,3584,0.04110577702522278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,3072,0.04011377692222595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,51200,0.3985315428839789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,2048,0.0369466675652398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,2560,0.03427911135885451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,6144,0.05996888875961304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,2560,0.02898400028546651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,3072,0.03722666700681051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,1536,0.028672890530692205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,2048,0.019533332851197984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,1024,0.019984000258975558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,2560,0.03413777881198459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,1536,0.01630311045381758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,1536,0.02313511073589325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,65536,0.5154204368591309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,768,0.014883556299739413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,2048,0.029846221208572388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,1024,0.01690933273898231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,1024,0.014783999986118741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,512,0.010564444793595208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,1536,0.024847999215126038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,768,0.01349066694577535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,1024,0.020198222663667467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,768,0.014124444789356656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,256,0.006400888992680444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,512,0.01108888867828581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,512,0.011215999722480774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,128,0.003144888828198115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,256,0.009364444348547194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,256,0.008014221986134848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,512,0.017243555850452848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,64,0.002554666664865282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,128,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,128,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,32,0.002711999954448806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,64,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,2048,0.0288702216413286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,128,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,65536,0.3215964370303684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,32,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,65536,0.17638933658599854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,51200,0.24918844964769152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,65536,0.15680178006490073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,768,0.019428445233239066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,51200,0.1307822200987074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,16384,0.08521066771613227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,256,0.0131217779384719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,16384,0.047803554270002574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,16384,0.050945778687795006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,51200,0.12053333388434516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,10240,0.0532968905236986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,16384,0.05409511261516147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,12288,0.03370755579736497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,12288,0.04862222075462341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,8192,0.04169955518510607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,12288,0.04565422071350945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,10240,0.029271112547980413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,10240,0.038041777080959745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,8192,0.024414221445719402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,8192,0.03494400117132399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,65536,0.1777777804268731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,7168,0.03657689028316074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,51200,0.1377857791052924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,6144,0.031898667414983116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,8192,0.03494133220778571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,7168,0.021931555536058214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,7168,0.030489779180950586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,5120,0.027435556054115295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,6144,0.02968533171547784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,6144,0.019705777366956074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,12288,0.0658551123407152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,4096,0.02276177704334259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,7168,0.031325333648257785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,5120,0.017458667357762653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,5120,0.024720888998773363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,3584,0.019832889238993328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,4096,0.01570399933391147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,4096,0.021187555458810594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,3072,0.017309332887331646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,3584,0.014589332871966891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,10240,0.041259557008743286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,3584,0.018980445133315194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,4096,0.022953778505325317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,2560,0.014843554960356818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,3072,0.01349688900841607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,3072,0.017229333519935608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,2048,0.012245333029164208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,3584,0.02085599965519375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,2560,0.015506666567590503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,2560,0.012381333443853589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,1536,0.010004444254769219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,3072,0.01961777773168352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,2048,0.010880889164076911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,2048,0.013381332986884646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,1024,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,2560,0.01794399983353085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,1536,0.009330666727489894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,2048,0.016819554898473952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,768,0.006231110956933763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,6144,0.02791200081507365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,1024,0.009665778113736046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,1536,0.015099555253982544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,512,0.004048888882001241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,1024,0.013953778478834363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,768,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,768,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,256,0.0024666666156715816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,512,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,512,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,768,0.012313777373896705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,5120,0.02553244431813558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,128,0.0021333334346612296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,256,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,256,0.00665155549844106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,64,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,512,0.011047999891969891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,32,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,128,0.006706666615274217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,128,0.005315555466545953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,256,0.011047999891969891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,64,0.006088888893524806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,128,0.008955555657545725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,65536,0.24442754851447213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,1536,0.01146577795346578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,65536,0.1484746668073866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,1024,0.008632889224423302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,51200,0.1892533302307129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,16384,0.06472711430655585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,65536,0.14038222365909153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,51200,0.10288888878292507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,16384,0.04105866617626614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,12288,0.05015022224850125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,16384,0.03737511237462362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,16384,0.044405334525638156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,12288,0.029919998513327703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,32,0.006174222048785951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,12288,0.04048000110520257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,10240,0.04252533449067009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,51200,0.10643466975953843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,12288,0.03586311141649882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,8192,0.03535288903448317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,10240,0.025817778375413682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,10240,0.03756622142261929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,7168,0.03092799915207757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,51200,0.1121564441257053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,8192,0.02181511123975118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,10240,0.032283554474512734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,8192,0.030653334326214258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,6144,0.027064000566800434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,7168,0.019873778025309246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,8192,0.029198222690158423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,7168,0.02734311090575324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,5120,0.023620444867346022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,6144,0.024359110328886244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,6144,0.017995556195576985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,7168,0.02644888891114129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,5120,0.016167110866970487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,6144,0.023988443944189284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,5120,0.0221413340833452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,3584,0.017887999614079792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,4096,0.014369777507252164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,4096,0.01960711181163788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,65536,0.13765422503153482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,3584,0.017539555827776592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,3072,0.015826667348543804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,3584,0.013264888690577613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,4096,0.0195448895295461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,2560,0.01347555551264021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,3072,0.012133333418104382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,3584,0.018367111682891846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,3072,0.015832889411184523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,2048,0.011723555624485016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,2560,0.011141333315107556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,2560,0.01421511173248291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,1536,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,3072,0.01755644381046295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,2048,0.009734222458468543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,2048,0.01242844429281023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,2560,0.016343111793200176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,1024,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,4096,0.02077155477470822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,1536,0.008446222378147973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,1536,0.010627555350462595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,768,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,2048,0.015565334094895257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,1024,0.0081395556529363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,1024,0.008051555189821456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,512,0.0031591111587153543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,768,0.0076248885856734375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,1536,0.014416888356208801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,256,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,5120,0.021864889396561518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,768,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,1024,0.011838222543398539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,128,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,512,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,512,0.00630666646692488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,256,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,64,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,256,0.005886222339338726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,512,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,32,0.0019706665641731685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,128,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,128,0.005442666510740916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,256,0.010815111299355825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,64,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,128,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,65536,0.2053368886311849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,32,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,51200,0.16155733002556696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,65536,0.13935022883945042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,16384,0.0573635564910041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,65536,0.12534488572014704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,51200,0.10675644212298924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,51200,0.10007555617226495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,16384,0.03533777925703261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,16384,0.039155556095971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,12288,0.02795555525355869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,12288,0.03558399942186143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,768,0.01090400003724628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,10240,0.03674222363366021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,51200,0.09340177641974555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,12288,0.03318311108483209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,65536,0.12022311157650417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,8192,0.03025777803526984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,10240,0.03169511093033685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,10240,0.024474667178259954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,7168,0.02732622292306688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,8192,0.02088622252146403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,10240,0.02960977620548672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,8192,0.026362667481104534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,6144,0.024255999260478552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,7168,0.019004444281260174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,7168,0.02347466680738661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,12288,0.04264622264438205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,6144,0.020616888999938965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,5120,0.020943111843532983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,6144,0.017276444368892245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,7168,0.024314666787783306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,4096,0.017509332961506315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,5120,0.015441776977645027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,6144,0.022509333160188463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,16384,0.040307554933759905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,3584,0.01552088889810774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,4096,0.01640266676743825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,5120,0.019951999187469482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,4096,0.013803555733627744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,3072,0.014003554979960123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,3584,0.01278133359220293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,4096,0.01828533411026001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,3584,0.015187554889255099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,2560,0.012447111308574677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,3072,0.013473778135246701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,3072,0.011734222372372946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,3584,0.018055111169815063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,2048,0.010336000058386061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,2560,0.010325333310498131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,3072,0.01664444473054674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,2048,0.011039999624093374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,2560,0.015968888998031616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,2048,0.009199110998047723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,5120,0.018573333819707234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,2048,0.014916444818178812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,1024,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,1536,0.009731555150614845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,1536,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,1536,0.013752889302041797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,768,0.0039822223285834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,1024,0.007069333559936947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,1024,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,768,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,512,0.0028088889602157804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,1024,0.0110880004035102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,768,0.006683555742104848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,256,0.0021964444054497615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,8192,0.027132444911532935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,768,0.010679999987284342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,512,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,128,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,256,0.005314666777849197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,256,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,1536,0.008265777594513362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,64,0.0018977777411540349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,128,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,32,0.0021182222084866627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,256,0.0107342220014996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,128,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,64,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,65536,0.1694257789187961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,128,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,32,0.005658666706747479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,51200,0.12725688351525202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,65536,0.08953955438401963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,65536,0.07053511010275947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,51200,0.06871733400556776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,16384,0.04356000158521864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,51200,0.05834488736258613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,512,0.006679111056857639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,12288,0.033957332372665405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,16384,0.027407111393080816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,512,0.010858666565683154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,12288,0.026760000321600173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,65536,0.11067110962337917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,10240,0.02663644486003452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,51200,0.08929422166612412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,12288,0.019186667270130582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,8192,0.02273155583275689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,10240,0.016415999995337594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,12288,0.03128000100453695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,10240,0.024030221833123103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,7168,0.0196106665664249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,8192,0.019944889677895438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,8192,0.014456889695591398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,10240,0.028207110034094915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,6144,0.01720888912677765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,7168,0.013464888764752282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,7168,0.018566222654448617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,8192,0.024960888756646052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,5120,0.01479111115137736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,6144,0.016543111867374845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,6144,0.012720888687504662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,7168,0.022847111026446026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,16384,0.02469066613250309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,4096,0.012506666282812754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,5120,0.012038222617573209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,5120,0.014454222387737699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,16384,0.03827022181616889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,3584,0.010975999964608086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,4096,0.013111111190583972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,4096,0.010418666733635796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,5120,0.019839111301634047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,3072,0.00963466697269016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,3584,0.009520000053776635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,4096,0.01816177699300978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,3072,0.011635555161370171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,2560,0.008428444465001425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,3072,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,3584,0.01718577742576599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,2560,0.008347555167145198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,2560,0.012340444657537671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,2048,0.007197333706749811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,2560,0.01036622209681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,2560,0.015556444724400839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,2048,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,1536,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,2048,0.009304888546466827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,6144,0.020980444219377305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,1024,0.004535111288229625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,1536,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,2048,0.014543111125628153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,1536,0.007607999775144789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,768,0.0034968890249729156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,1024,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,1024,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,1536,0.013426666458447775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,512,0.0028248888750871024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,3584,0.012426666915416718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,768,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,768,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,256,0.0023093333260880574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,1024,0.011761777930789523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,512,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,768,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,128,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,3072,0.016234666109085083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,256,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,256,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,512,0.011219555305110084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,64,0.0019537777536445195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,32,0.0018595556418100994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,128,0.005636444522274866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,256,0.010679999987284342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,128,0.004936888813972473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,64,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,65536,0.14755733807881674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,32,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,128,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,51200,0.1163431141111586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,65536,0.08283378018273248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,16384,0.039800001515282526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,65536,0.065647111998664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,51200,0.06237600247065226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,51200,0.053471999035941235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,16384,0.02588533361752828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,16384,0.021732444564501446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,512,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,12288,0.01789422167672051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,10240,0.026996443669001263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,65536,0.10956888728671604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,12288,0.025751110580232408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,51200,0.08899377452002631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,8192,0.02277511192692651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,10240,0.015608888533380298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,12288,0.03146133489078946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,10240,0.022652443912294176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,7168,0.0177208897140291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,8192,0.01958311100800832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,8192,0.01404622197151184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,6144,0.015442666080262927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,10240,0.028034667174021404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,7168,0.017549332645204332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,8192,0.02444444431198968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,7168,0.013024888932704926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,5120,0.013340444200568728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,6144,0.011964444484975604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,6144,0.0163128889269299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,7168,0.023043556345833674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,4096,0.011348444554540845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,5120,0.014121777481502958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,12288,0.031664000617133245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,5120,0.011058666639857821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,3584,0.010129777921570672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,6144,0.021164443757798936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,4096,0.009639999932712978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,5120,0.01921066641807556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,4096,0.012412444584899478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,3072,0.00904977818330129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,3584,0.009038222332795462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,2560,0.008014221986134848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,3584,0.011712888876597086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,16384,0.03780000077353583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,4096,0.017708443933063082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,3072,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,2048,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,3072,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,2560,0.007659555309348636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,3584,0.016928889685206942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,1536,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,2560,0.010034666293197209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,3072,0.01681688924630483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,2048,0.007141333487298753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,2048,0.008683555656009251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,1024,0.004074666649103165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,2560,0.015708444847000968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,1536,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,1536,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,768,0.0032257777121331957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,1024,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,1536,0.013156444662147097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,512,0.0025555555605226094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,768,0.006226666685607698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,768,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,1024,0.011008000208271874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,256,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,512,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,512,0.00564533347884814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,768,0.011374221907721626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,128,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,256,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,256,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,512,0.010055999788973067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,64,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,128,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,256,0.010052444206343757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,128,0.004947555561860402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,65536,0.12079732947879368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,128,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,32,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,51200,0.09506133529875015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,65536,0.07407910956276788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,2048,0.014552889598740472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,65536,0.0635111133257548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,51200,0.05638844437069363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,16384,0.03397066725624932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,51200,0.05155555407206217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,12288,0.026854222019513447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,16384,0.021368889345063105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,16384,0.023456000619464453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,51200,0.08897599909040664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,10240,0.022964444425370958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,12288,0.01702844434314304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,12288,0.02311466634273529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,10240,0.015205333630243937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,12288,0.030961778428819444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,10240,0.02068444424205356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,8192,0.01982666717635261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,64,0.005283555636803309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,7168,0.01755555636352963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,10240,0.027356444133652583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,8192,0.013647110925780402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,6144,0.015506666567590503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,8192,0.018182223041852314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,7168,0.016286222471131217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,7168,0.01254755589697096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,5120,0.013607111242082385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,8192,0.02383822202682495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,6144,0.011514667007658215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,7168,0.022310222188631695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,6144,0.014946666028764514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,4096,0.010924444430404238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,5120,0.013829333086808523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,65536,0.10999555720223321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,3584,0.009333333207501305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,6144,0.020631111330456205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,4096,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,4096,0.008959111240175035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,16384,0.03695377707481384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,5120,0.018928888771269057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,3584,0.008661333057615492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,4096,0.01791111131509145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,3584,0.011400000088744693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,2560,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,3072,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,3584,0.016964443855815463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,3072,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,2048,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,2560,0.007327111230956183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,2560,0.009283555878533257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,3072,0.01603555513752831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,1536,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,2048,0.007735999921957652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,2048,0.006948444578382704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,1024,0.003549333247873518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,1536,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,2048,0.013337777720557319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,1536,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,768,0.002830222249031067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,1024,0.006311999840868845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,5120,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,3072,0.008268444074524773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,1024,0.011001778145631155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,768,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,512,0.002456888970401552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,768,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,768,0.01038933296998342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,256,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,512,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,512,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,256,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,512,0.010327999790509542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,2560,0.015496888094478183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,256,0.009673777553770278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,128,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,128,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,1536,0.011996444728639392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,1024,0.006351999938488007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,32,0.0018719999740521114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,128,0.008700444466537898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,65536,0.1048097809155782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,64,0.005139555368158552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,32,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,65536,0.06735644737879436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,51200,0.08282933632532756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,65536,0.06116444534725613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,51200,0.05103288756476509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,16384,0.030459556314680312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,128,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,256,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,65536,0.10983288950390285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,12288,0.024433778391944036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,16384,0.02015911042690277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,51200,0.08806400166617499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,16384,0.021292444732454088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,64,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,10240,0.020530665914217632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,12288,0.01681777834892273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,16384,0.03672800130314297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,12288,0.022614222433831956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,8192,0.01814133259985182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,10240,0.014921777778201632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,10240,0.02018311123053233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,7168,0.016033778587977093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,8192,0.017853332890404593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,8192,0.01311288856797748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,6144,0.014085332552591959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,7168,0.012378666963842181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,8192,0.023419555690553453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,7168,0.015961777832772996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,5120,0.012272888587580787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,6144,0.014461333552996317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,6144,0.01112711098459032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,51200,0.049456000328063965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,7168,0.02264533274703556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,5120,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,5120,0.013474666410022311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,6144,0.02036355601416694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,4096,0.011700444751315646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,3584,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,5120,0.01885422236389584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,3584,0.008279110822412703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,4096,0.01754666699303521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,3072,0.0075395554304122925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,3584,0.010910222099887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,12288,0.030223111311594646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,2560,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,3584,0.01670488880740272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,10240,0.027249778310457867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,3072,0.010044444766309526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,2048,0.005759999983840519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,2560,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,2560,0.008355555435021719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,3072,0.01572711103492313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,1536,0.004524444540341695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,2048,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,4096,0.00963466697269016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,2560,0.014840889308187695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,1024,0.0032151111712058387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,1536,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,1536,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,2048,0.013497778111033969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,768,0.0028479999552170434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,4096,0.008621333373917473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,1024,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,1024,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,1536,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,512,0.0024577778660588795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,768,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,768,0.005789333333571752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,256,0.0021475555582178966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,1024,0.011046222514576383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,512,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,768,0.010682666467295753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,3072,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,128,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,512,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,256,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,64,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,128,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,256,0.00980444418059455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,128,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,2048,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,65536,0.09011644124984741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,128,0.008298666940795051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,64,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,32,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,51200,0.06542311112085979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,65536,0.05881422095828586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,65536,0.059874667061699756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,16384,0.02497600018978119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,51200,0.04653866754637825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,51200,0.04747466577423943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,65536,0.10953244235780503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,12288,0.019204444355434842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,512,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,16384,0.018907555275493197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,16384,0.019724445210562814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,51200,0.08808622095319961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,10240,0.01679644485314687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,256,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,12288,0.016146666473812528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,8192,0.014395554860432943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,12288,0.02142222225666046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,10240,0.014458666245142618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,10240,0.018026666508780587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,12288,0.030082666211658057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,7168,0.012793777717484368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,8192,0.012720000412729053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,8192,0.016819554898473952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,10240,0.026424888107511733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,6144,0.011776888536082374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,7168,0.011581333147154914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,8192,0.02368533280160692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,7168,0.014507555299335055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,5120,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,6144,0.014098667436175875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,7168,0.021791110436121624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,4096,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,5120,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,5120,0.01218933363755544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,6144,0.020208888583713107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,3584,0.011418666276666852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,5120,0.019009777241282992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,4096,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,4096,0.010654222634103564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,3072,0.008971555365456475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,3584,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,16384,0.036722666687435575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,4096,0.017095110482639737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,2560,0.008261333737108443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,3072,0.008389333056079017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,3072,0.007618666523032718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,2560,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,2560,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,3072,0.015827554795477126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,2048,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,1536,0.005537777725193236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,2048,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,6144,0.010425777898894416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,2048,0.007336888876226213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,1024,0.004190222256713443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,1536,0.006322666588756774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,1536,0.006692444284756978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,2048,0.013111111190583972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,768,0.003546666767862108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,1024,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,1024,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,1536,0.011708444191349877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,512,0.0031742221779293488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,3584,0.010392000277837118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,768,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,768,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,1024,0.011121778024567498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,256,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,512,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,512,0.005399999933110342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,3584,0.016403555870056152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,128,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,256,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,256,0.005712888720962737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,64,0.0018800000349680583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,512,0.010040889183680216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,32,0.0018862223045693503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,128,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,128,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,2560,0.013725333743625216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,65536,0.07771022452248467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,128,0.008319999608728621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,32,0.00509688879052798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,51200,0.05968533621893989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,65536,0.05759377612007988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,65536,0.057877335283491343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,16384,0.022864888111750286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,51200,0.044799112611346774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,51200,0.046668443414900035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,65536,0.10947466558880276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,16384,0.01924622224436866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,16384,0.01827822294500139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,51200,0.08749066458808051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,768,0.010686222049925061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,16384,0.03565333286921183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,10240,0.015393777026070489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,12288,0.015798222687509324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,256,0.0103502223889033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,12288,0.029457777738571167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,8192,0.01295111080010732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,64,0.0052915554907586836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,10240,0.017673777209387887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,10240,0.014310222533014087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,7168,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,8192,0.012105777859687805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,8192,0.015219555960761176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,7168,0.0144213338692983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,6144,0.011062222222487131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,7168,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,8192,0.02390044430891673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,12288,0.017525333497259352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,6144,0.010245333115259806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,5120,0.009813333551088968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,6144,0.013072888884279462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,5120,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,6144,0.020198222663667467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,12288,0.019176888797018263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,5120,0.011779555843936073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,5120,0.0189057770702574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,4096,0.008332444561852349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,3584,0.010012444522645738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,4096,0.010227555202113258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,3584,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,10240,0.026400888959566753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,3584,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,3072,0.008743111458089616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,3584,0.015498666299713982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,2560,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,3072,0.007853333320882585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,7168,0.0218240006102456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,3072,0.015159999330838522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,2048,0.006829333388143116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,2560,0.007154666715198093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,2560,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,4096,0.008692444198661381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,2560,0.013408888545301227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,1536,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,2048,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,1536,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,1024,0.003932444585694207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,2048,0.01273333364062839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,1536,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,768,0.0034799998005231223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,4096,0.017351999878883362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,1024,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,1536,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,512,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,1024,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,768,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,768,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,3072,0.007521777517265744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,256,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,1024,0.01108888867828581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,512,0.005817777580685086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,512,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,768,0.010728889041476779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,256,0.005047111047638787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,512,0.01037599974208408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,64,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,128,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,128,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,256,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,32,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,65536,0.0620959997177124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,2048,0.007080000307824876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,128,0.00832444429397583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,64,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,51200,0.04866933491494921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,32,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,65536,0.055065777566697865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,65536,0.05714311202367147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,16384,0.01899377836121453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,51200,0.04311022162437439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,16384,0.01754577789041731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,16384,0.018923555811246235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,12288,0.015589333242840238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,51200,0.08829955259958903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,16384,0.0354017780886756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,256,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,12288,0.018239110708236694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,12288,0.03001333276430766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,10240,0.016418667303191293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,8192,0.012030222349696688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,10240,0.013896889156765409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,10240,0.026142223013771906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,8192,0.0144177774588267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,7168,0.010997333460383944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,8192,0.011744889120260874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,51200,0.045851555135515004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,8192,0.023736000061035156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,65536,0.10956977473364936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,6144,0.009658666948477427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,7168,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,7168,0.012995555169052549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,128,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,6144,0.009663110805882348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,10240,0.01347733371787601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,6144,0.011471110913488599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,12288,0.015593777100245157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,4096,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,6144,0.020593777298927307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,5120,0.009012444151772393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,5120,0.010431999961535135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,3584,0.007538666327794393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,4096,0.00832799987660514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,5120,0.01829866733815935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,3584,0.007620444728268518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,3584,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,4096,0.017231111725171406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,3584,0.01607022186120351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,3072,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,2560,0.007522666619883643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,3072,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,3072,0.014544000228246054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,7168,0.021583999196688335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,5120,0.00905600024594201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,2048,0.006162666612201267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,2560,0.007258666886223688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,2560,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,1536,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,2048,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,2048,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,4096,0.008664888640244802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,1536,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,2048,0.012780444489585029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,1024,0.004294222427739037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,768,0.0035573335157500375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,1024,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,3072,0.008583111067612966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,1024,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,512,0.0028817777832349143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,768,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,768,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,1024,0.011054221954610614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,256,0.002512000087234709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,512,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,512,0.005231110999981563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,768,0.011058666639857821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,128,0.0022248888595236675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,256,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,256,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,512,0.01033333357837465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,64,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,128,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,128,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,32,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,256,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,65536,0.055495109823015004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,64,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,1536,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,128,0.008275555239783393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,51200,0.043140444490644664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,65536,0.05233866638607449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,1536,0.012078222301271228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,65536,0.05596799982918633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,16384,0.017630222770902846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,51200,0.041511999236212835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,51200,0.04522488845719231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,12288,0.014010666145218743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,65536,0.10904710822635227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,16384,0.017125333348910015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,16384,0.01886133352915446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,51200,0.08814666668574016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,12288,0.015037332971890768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,16384,0.03535022338231405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,12288,0.016569778323173523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,8192,0.010401777923107147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,10240,0.014817777607176038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,10240,0.013342222405804528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,12288,0.02942755487230089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,2560,0.013775111072593264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,7168,0.010057777994208867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,8192,0.010963555839326648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,10240,0.0262071109480328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,8192,0.013010666602187686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,6144,0.008649778034951951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,7168,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,7168,0.01203644441233741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,8192,0.023620444867346022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,5120,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,6144,0.009685333404276106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,6144,0.011032000184059143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,4096,0.007632888853549957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,5120,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,6144,0.019905777441130746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,5120,0.009783999787436591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,3584,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,4096,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,4096,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,3584,0.007885333564546373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,3584,0.007615110940403408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,3072,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,4096,0.01613066593805949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,3584,0.01499466598033905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,3072,0.007276444799370236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,2560,0.006746666712893381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,3072,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,32,0.004952888935804367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,3072,0.014418666561444601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,2048,0.006268444574541516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,2560,0.007197333706749811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,7168,0.02162844439347585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,2560,0.006649777707126405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,1536,0.005039999882380168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,2048,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,2048,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,2560,0.013423110875818463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,1024,0.0038622220357259116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,10240,0.012039110892348819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,1536,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,5120,0.018900444110234577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,768,0.0035422220826148987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,2048,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,1024,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,1024,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,1536,0.012069332930776807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,768,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,1024,0.011347555451922946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,768,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,512,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,256,0.002540444541308615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,512,0.005787555542257097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,256,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,512,0.010325333310498131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,256,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,256,0.009376000199053023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,128,0.004871110949251386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,64,0.001790222194459703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,128,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,1536,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,32,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,128,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,65536,0.04856444398562113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,64,0.004861333303981357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,512,0.0028844444702068963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,32,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,51200,0.03687377770741781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,768,0.005492444253630108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,65536,0.04974488748444451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,65536,0.05532177620463901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,16384,0.014511111709806653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,51200,0.04455555478731791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,51200,0.039992888768514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,12288,0.011735999749766456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,65536,0.10895466804504395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,16384,0.0185217774576611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,16384,0.016787555482652452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,10240,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,128,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,12288,0.0144533332851198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,51200,0.08787733316421509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,8192,0.009570666485362584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,12288,0.01590666671593984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,16384,0.03552444444762336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,10240,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,10240,0.014315555493036905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,12288,0.02964444292916192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,8192,0.010695111420419482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,8192,0.011994666523403592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,7168,0.010087110930018956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,7168,0.010678222609890832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,8192,0.023003555006451074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,5120,0.007668444679843054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,6144,0.008801778157552084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,6144,0.009674666656388177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,7168,0.021184888150956895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,4096,0.009036444127559662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,5120,0.008978666530715095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,5120,0.008632889224423302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,6144,0.020251555575264823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,3584,0.008244444098737504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,4096,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,4096,0.007560888926188151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,5120,0.017246221502621967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,3072,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,3584,0.007827555967701806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,4096,0.015798222687509324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,7168,0.008840888738632202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,2560,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,3072,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,3584,0.01494311127397749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,3072,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,6144,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,2048,0.006248888870080312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,10240,0.02676444417900509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,2560,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,3072,0.01440622243616316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,2560,0.006659555352396435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,1536,0.005512889060709212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,2048,0.006706666615274217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,2048,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,1024,0.003847111016511917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,1536,0.006574222197135289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,2560,0.013448889056841532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,1536,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,2048,0.012780444489585029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,768,0.003527111063400904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,1024,0.006265777680608962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,1536,0.011786667009194693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,512,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,768,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,768,0.005384888913896348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,1024,0.010777778095669217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,256,0.002479111154874166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,3584,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,768,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,512,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,512,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,128,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,256,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,256,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,512,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,64,0.0019128889673286015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,32,0.0018880000958840053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,128,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,256,0.010406222608354356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,65536,0.03413866625891791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,128,0.008012444608741337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,64,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,51200,0.027122666438420612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,32,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,65536,0.054735110865698926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,65536,0.04854311214552986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,16384,0.012070222033394707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,51200,0.04443822304407755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,51200,0.03898755709330241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,1024,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,12288,0.009676444033781687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,65536,0.1091724435488383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,16384,0.016183111402723525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,10240,0.008716444174448649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,51200,0.08775822321573894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,12288,0.013376889129479727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,16384,0.03529777791765001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,12288,0.014592889282438489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,8192,0.0075333333677715724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,10240,0.011915555430783166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,10240,0.012104888757069906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,12288,0.02878755662176344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,8192,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,8192,0.010423999693658618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,10240,0.025655110677083332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,6144,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,8192,0.022488888767030504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,7168,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,7168,0.009335999687512716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,5120,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,6144,0.009321778184837764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,7168,0.021685333715544805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,6144,0.008700444466537898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,4096,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,5120,0.008303110798199972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,5120,0.008661333057615492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,6144,0.01923111081123352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,3584,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,16384,0.01723288827472263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,4096,0.008017777568764156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,4096,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,5120,0.01755822201569875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,3072,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,3584,0.00777333312564426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,2560,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,3584,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,3072,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,3072,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,3584,0.014836443795098199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,2048,0.005470222069157495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,2560,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,2560,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,1536,0.004634666774008009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,3072,0.01443733274936676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,2048,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,2560,0.01333866682317522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,2048,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,1536,0.0063928887248039246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,1536,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,2048,0.012727999852763282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,768,0.003169777699642711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,1024,0.005543111099137201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,1536,0.012115555504957834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,768,0.005974222388532426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,768,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,1024,0.011622221933470832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,768,0.010666666759385003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,4096,0.015765332513385348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,512,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,256,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,7168,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,512,0.005537777725193236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,128,0.0021333334346612296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,512,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,256,0.005591111050711737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,256,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,1024,0.003563555578390757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,64,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,256,0.010061333576838175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,128,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,32,0.0018684443914228016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,65536,0.024081778195169237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,128,0.008375999828179678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,64,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,32,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,51200,0.019895111521085102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,65536,0.04658044377962748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,65536,0.054436445236206055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,51200,0.043861332866880626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,51200,0.037540445725123085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,16384,0.013117333253224691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,65536,0.10849333471722072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,51200,0.0865066647529602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,16384,0.014831999937693277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,12288,0.010461333725187512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,16384,0.03434755404790243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,128,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,10240,0.009663999908500248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,12288,0.01237422227859497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,12288,0.013073777986897362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,1024,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,8192,0.009087111386987898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,12288,0.028368890285491943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,10240,0.011214222345087262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,7168,0.008991111483838823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,512,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,8192,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,8192,0.009744000103738572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,6144,0.007183111376232571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,7168,0.00998755544424057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,8192,0.02165777815712823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,7168,0.020592888196309406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,6144,0.009402666654851701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,16384,0.015983111328548856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,6144,0.018740443719757926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,5120,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,4096,0.005867555737495422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,5120,0.008626666333940294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,5120,0.01719555589887831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,10240,0.011758222348160215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,4096,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,3584,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,4096,0.015488889482286243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,10240,0.025090666280852422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,3072,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,7168,0.009347555538018545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,3584,0.007328888608349695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,3584,0.007584888901975419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,2560,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,3584,0.014557333456145393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,3072,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,3072,0.006945778098371293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,5120,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,2048,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,2560,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,2560,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,1536,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,3072,0.014473777678277759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,2048,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,2048,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,2560,0.012853333519564735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,1024,0.0038559999730851916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,1536,0.006112888869312074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,2048,0.012747555143303342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,1536,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,768,0.003359111025929451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,1024,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,1024,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,1536,0.012075555821259817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,512,0.0028044444819291434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,768,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,768,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,256,0.0022453332526816265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,512,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,768,0.010339555641015371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,256,0.004951111144489712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,256,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,128,0.0021786666992637846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,6144,0.0087013335691558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,128,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,256,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,32,0.0023048888478014204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,128,0.0047253333032131195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,65536,0.019172444939613342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,128,0.008339555727110969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,64,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,32,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,51200,0.015863999724388123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,65536,0.045893334680133395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,65536,0.05413155423270332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,1024,0.011415111521879831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,16384,0.013463111387358772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,51200,0.03700266612900628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,65536,0.10818044344584148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,512,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,512,0.009999111294746399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,16384,0.015841777125994366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,51200,0.08649866448508368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,12288,0.010037333601050908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,16384,0.01460800071557363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,10240,0.008371555142932469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,12288,0.013055110971132914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,16384,0.03460888730155097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,12288,0.012420444852775998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,8192,0.007768889268239339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,10240,0.010991999672518836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,10240,0.011765333513418833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,12288,0.027979556057188246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,7168,0.007028444773621029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,8192,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,8192,0.009815110928482479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,10240,0.024671110841963027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,6144,0.006337777607970768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,7168,0.009341333475377824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,7168,0.010041777458455827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,8192,0.021670222282409668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,5120,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,51200,0.04309066798951891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,6144,0.009338666995366415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,6144,0.00868711123863856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,7168,0.020608888732062448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,4096,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,5120,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,5120,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,3584,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,6144,0.018242667118708294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,4096,0.008178667061858708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,4096,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,5120,0.01722133325205909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,3072,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,3584,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,3584,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,2560,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,3072,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,3584,0.01479466590616438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,3072,0.014119999276267158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,2560,0.006607111129495833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,4096,0.007546666595670912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,2048,0.0042133331298828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,2560,0.006913777854707506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,1536,0.004308444344335132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,2048,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,2560,0.013816888961527081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,2048,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,1024,0.00349866681628757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,1536,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,1536,0.006320000108745363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,2048,0.012846222354306115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,768,0.0031991110493739447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,1024,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,1024,0.005535999933878581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,512,0.002532444480392668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,768,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,768,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,4096,0.015171556009186639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,256,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,768,0.011027555498811932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,512,0.0051857779423395796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,3072,0.0069004446268081665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,128,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,256,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,256,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,512,0.010703999963071613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,64,0.0018666666001081467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,128,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,128,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,768,32,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,256,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,65536,0.02811288833618164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,64,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,128,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,32,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,51200,0.022358222140206233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,65536,0.04533955454826355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,65536,0.0516480008761088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,16384,0.00944622192117903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,1536,0.011768889096048145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,65536,0.10762933227750991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,51200,0.036455111371146307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,12288,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,1024,0.010417777631017896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,16384,0.014495111174053617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,16384,0.015825778245925903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,51200,0.08634044726689656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,10240,0.006904888898134232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,16384,0.03427911135885451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,12288,0.012055111428101858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,12288,0.013395555317401886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,8192,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,10240,0.011736888852384357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,10240,0.01089244418674045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,12288,0.0277502222193612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,7168,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,8192,0.009748444788985783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,8192,0.02157955533928341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,7168,0.009174221919642555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,6144,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,7168,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,7168,0.020602666669421725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,51200,0.038683556848102145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,6144,0.00868622213602066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,5120,0.005504888792832692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,6144,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,512,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,4096,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,5120,0.00871022211180793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,5120,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,4096,0.007974222302436829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,5120,0.01756088932355245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,3584,0.0053635554181204895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,4096,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,10240,0.02501511077086131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,3072,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,3584,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,3584,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,3584,0.014527999692493014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,3072,0.007371555599901412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,3072,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,2560,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,3072,0.014409777190950183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,2560,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,2048,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,2560,0.006329777754015393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,6144,0.018534221582942538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,2560,0.013431111143694984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,2048,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,2048,0.012711111042234631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,1536,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,1024,0.00309244439833694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,8192,0.010401777923107147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,4096,0.016161777906947665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,1536,0.011514667007658215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,768,0.0028213332924577924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,1024,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,1024,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,512,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,1024,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,768,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,512,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,256,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,2048,0.006934222247865465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,1536,0.0035946667194366455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,512,0.005655111124118169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,512,0.01038755559258991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,128,0.0018888889915413326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,256,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,256,0.004608889006906086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,64,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,128,0.005453333258628845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,128,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,512,32,0.0017804443422291016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,256,0.01036088913679123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,65536,0.016509332590632968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,128,0.008649778034951951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,64,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,51200,0.014103111293580798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,32,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,65536,0.04291022155019972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,768,0.0057751110030545135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,768,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,16384,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,51200,0.034097777472601995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,51200,0.03702400128046671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,65536,0.10691377851698135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,12288,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,16384,0.014136888914638095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,51200,0.08602755599551731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,10240,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,12288,0.012038222617573209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,12288,0.013111111190583972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,10240,0.011719110939237805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,12288,0.028210666444566514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,10240,0.010781333678298526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,8192,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,1536,0.006568888823191325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,7168,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,8192,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,10240,0.024680889315075342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,8192,0.00978311151266098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,6144,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,7168,0.009324444664849175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,65536,0.046441776884926685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,6144,0.008755555583371056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,5120,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,7168,0.019948444432682462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,16384,0.01551822159025404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,6144,0.018573333819707234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,4096,0.004509333521127701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,5120,0.00830666638082928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,16384,0.03388444582621256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,5120,0.016650666793187458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,4096,0.007925333248244392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,3584,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,4096,0.0075457774930530125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,4096,0.016155555844306946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,3584,0.006751110984219446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,3072,0.004575110971927643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,3584,0.007788444558779399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,7168,0.009698666632175446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,8192,0.021969777014520433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,3584,0.01481599940194024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,2560,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,3072,0.007248000138335758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,3072,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,2048,0.004557333472702238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,2560,0.006911110960774952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,2560,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,3072,0.01443288889196184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,1536,0.0035617777870761026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,2560,0.013455111119482251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,2048,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,2048,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,5120,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,1024,0.002846222163902389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,1536,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,2048,0.012105777859687805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,1536,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,768,0.002537777854336633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,1024,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,1024,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,512,0.00254844439526399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,1536,0.011729777687125735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,1024,0.011393778026103973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,768,0.0057795556883017225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,768,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,512,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,512,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,768,0.010346666806273991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,128,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,256,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,256,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,512,0.01036977767944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,64,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,32,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,128,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,128,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,65536,0.011399110986126794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,256,0.00999022192425198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,64,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,51200,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,32,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,128,0.00869511150651508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,65536,0.04586844311820137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,16384,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,65536,0.041950222518708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,51200,0.03698933455679152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,51200,0.033546666304270424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,12288,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,65536,0.10684000121222602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,16384,0.014382221632533602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,16384,0.015798222687509324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,10240,0.0052879999081293745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,51200,0.08533155255847508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,12288,0.013096888860066732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,16384,0.03387733300526937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,12288,0.011745777395036487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,8192,0.005201777650250329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,10240,0.011034666664070554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,6144,0.009352000223265754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,7168,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,8192,0.009687110781669617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,10240,0.02504355543189579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,8192,0.010683555569913654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,6144,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,8192,0.021594666772418555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,7168,0.009687110781669617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,7168,0.009333333207501305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,5120,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,6144,0.009188444250159794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,6144,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,7168,0.019928889142142404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,4096,0.0041831110914548235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,5120,0.008748444418112436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,5120,0.00812355594502555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,6144,0.0185199992524253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,3584,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,4096,0.008019555773999956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,256,256,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,4096,0.007359999749395583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,5120,0.01719377769364251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,3072,0.0042275554604000514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,3584,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,3584,0.007449777589903937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,4096,0.015838222371207345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,10240,0.011755555868148804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,2560,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,12288,0.027446221974160936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,3072,0.0070053330726093715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,2048,0.003955555458863576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,3584,0.015160888433456421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,2560,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,1536,0.0034835553831524322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,3072,0.013754666679435305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,2560,0.00647288892004225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,2048,0.006763555523422029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,2048,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,1024,0.0028728888266616394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,2560,0.01331022216214074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,1536,0.006351999938488007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,2048,0.01237511138121287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,1536,0.005778666585683823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,768,0.0025804444319672054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,1024,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,1024,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,1536,0.012087999946541257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,512,0.002508444504605399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,768,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,768,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,256,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,1024,0.011394666300879585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,512,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,768,0.010378667049937779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,512,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,128,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,256,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,256,0.0047048889100551605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,512,0.010295111272070143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,64,0.0018071112119489247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,128,32,0.001791999985774358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,3072,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,65536,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,128,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,256,0.009712888962692684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,51200,0.006320000108745363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,16384,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,64,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,128,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,12288,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,32,0.004594666676388847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,10240,0.004567111117972268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,8192,0.004613333278232151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,65536,0.041469332244661115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,51200,0.03360177742110358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,7168,0.004550222307443619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,16384,0.014434667097197639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,6144,0.004268444246715969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,5120,0.004588444613748127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,12288,0.011751111182901593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,10240,0.0107360002067354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,4096,0.0038853333228164246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,3584,0.004271111140648524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,7168,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,6144,0.00871022211180793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,5120,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,4096,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,3584,0.006686222222116258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,2048,0.00424177779091729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,3072,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,1536,0.0035333335399627686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,2560,0.006457777900828256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,1024,0.0028044444819291434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,2048,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,768,0.002524444419476721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,1536,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,512,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,1024,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,256,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,768,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,128,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,64,0.0017991111510329777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,512,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,32,0.0018168888572189545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,256,0.004595555365085602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,8192,0.009530666801664565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,65536,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,51200,0.006565333240562015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,128,0.00453333349691497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,3072,0.0042133331298828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,16384,0.005118222286303838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,64,0.004591111093759537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,12288,0.004595555365085602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,32,0.005031110925806893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,64,2560,0.004535111288229625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,10240,0.004582222137186262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,8192,0.004193777839342753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,65536,0.04178222351604038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,7168,0.004252444538805219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,6144,0.004264889078007804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,51200,0.033569776349597506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,16384,0.01443199978934394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,5120,0.004212444441186057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,12288,0.0120328888297081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,10240,0.010869333313571082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,8192,0.009371555513805812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,3584,0.004567999806669024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,7168,0.009032888544930352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,3072,0.004524444540341695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,128,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,6144,0.008365333080291748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,2048,0.003919110943873723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,5120,0.00796622203456031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,4096,0.006986666884687211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,1536,0.003506666670242945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,3584,0.006649777707126405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,1024,0.003028444531891081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,3072,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,2560,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,768,0.002540444541308615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,512,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,2048,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,256,0.0018586667461527719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,1536,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,128,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,1024,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,64,0.0017813332378864288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,768,0.005330666485759947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,32,0.0018044445249769424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,512,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,256,0.004538666870858935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,128,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,64,0.004593777573770947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,32,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,4096,0.0038311113086011675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,8,32,2560,0.004615111069546806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,16384,0.3376364443037245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,16384,0.21347999572753906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,51200,0.4868062337239583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,12288,0.2573244571685791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,16384,0.13509955671098497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,51200,1.0143831041124132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,10240,0.20123467180464003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,12288,0.15299288431803384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,12288,0.12231466505262588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,10240,0.13104444079928929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,10240,0.08677777979109023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,8192,0.07071644730038114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,7168,0.1319813330968221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,8192,0.11984977457258437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,6144,0.11378755834367539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,7168,0.09647733635372585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,5120,0.09590666823916966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,51200,0.40832000308566624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,4096,0.08067822456359863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,6144,0.08709777726067437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,6144,0.05385511120160421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,5120,0.04576177729500664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,3584,0.06804088751475017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,4096,0.039224889543321394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,4096,0.06804177496168348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,3584,0.03451555636194017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,3584,0.053736888700061373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,2560,0.050909333758884005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,3072,0.04063377777735392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,3072,0.04618577824698555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,8192,0.15337955951690674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,2560,0.027032888597912256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,2048,0.04169244567553202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,1536,0.032292445500691734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,2560,0.03912622067663405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,1024,0.022662222385406494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,2048,0.02299555473857456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,1536,0.018739556272824604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,2048,0.036293334431118436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,768,0.017830222845077515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,1536,0.028766221470303003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,7168,0.06264711088604398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,512,0.012702222499582501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,1024,0.020220445262061227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,1024,0.017861333158281114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,256,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,768,0.014846222268210517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,128,0.005126222140259213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,512,0.01275022245115704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,5120,0.07435111204783122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,256,0.010423111418883005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,512,0.012999999854299756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,64,0.004187555362780889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,32,0.0041759999261962045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,256,0.010495999621020423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,3072,0.05944710969924927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,768,0.017051556044154696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,128,0.010688888529936472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,128,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,64,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,32,0.009327111144860586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,16384,0.2495573361714681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,16384,0.1855644385019938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,51200,0.38947200775146484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,65536,0.5144631067911783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,51200,0.7471671104431152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,12288,0.19103377395206025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,65536,0.9641048643324109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,10240,0.15916889243655735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,16384,0.10629155238469441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,12288,0.11926666895548503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,12288,0.08369511365890503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,8192,0.13891200224558511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,51200,0.36207111676534015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,8192,0.09215555588404338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,10240,0.0999893347422282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,8192,0.0596311092376709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,7168,0.12173155943552653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,7168,0.049787554475996226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,10240,0.07161066929499309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,6144,0.10480622450510661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,65536,0.4276417891184489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,6144,0.04575555523236593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,5120,0.08853511015574138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,6144,0.07123021947013007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,4096,0.07179288731680976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,7168,0.07879822121726142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,5120,0.05948088566462199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,4096,0.0317555566628774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,3072,0.05420710974269443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,4096,0.05471733212471008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,3584,0.029166221618652344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,2560,0.045534223318099976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,3584,0.04451466600100199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,3072,0.026353778110610113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,3072,0.03815111186769273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,2560,0.02678666677739885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,2560,0.032624890406926475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,2048,0.01902311046918233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,2048,0.029225776592890423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,1024,0.020224889119466145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,1536,0.01608888970481025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,1536,0.02327377763059404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,768,0.01497066683239407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,1024,0.015022221538755627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,1024,0.01735822194152408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,512,0.010498666928874122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,768,0.013201778133710226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,3584,0.06291644440756904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,5120,0.03793599870469835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,256,0.006321777900060018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,512,0.011046222514576383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,128,0.0034097778714365433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,256,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,2048,0.03678399986690945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,64,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,128,0.009307555854320526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,32,0.002817777709828483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,1536,0.0284862220287323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,64,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,32,0.008416000339719985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,65536,0.32256356875101727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,65536,0.17173243893517387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,51200,0.24869155883789062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,768,0.01444355481200748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,65536,0.15829688972897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,16384,0.08482933044433594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,256,0.0085235552655326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,51200,0.1306568913989597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,128,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,12288,0.06549600097868177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,16384,0.06776889165242513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,10240,0.05308622121810913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,16384,0.0418488879998525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,51200,0.11934044626024033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,12288,0.03227555420663621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,12288,0.04873422119352552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,8192,0.04232711262173123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,7168,0.03663377629386054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,10240,0.04068622324201796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,6144,0.031713777118259005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,10240,0.028628445333904688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,7168,0.021912000245518152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,8192,0.03683288892110189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,7168,0.03143289022975498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,5120,0.027303111222055223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,6144,0.019655111763212416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,4096,0.022711111439598933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,3584,0.01995111174053616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,5120,0.017507554756270517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,4096,0.015400000744395785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,512,0.011191111471917896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,3072,0.017147555947303772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,4096,0.021707556313938562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,2560,0.014924445086055331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,3584,0.014255111416180929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,5120,0.024596444434589807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,3584,0.018573333819707234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,3072,0.013247999880048962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,2048,0.012143111063374413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,1536,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,3072,0.016252444850073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,2560,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,1024,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,2048,0.010804444551467896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,2560,0.014660444524553088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,768,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,1536,0.009012444151772393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,2048,0.013751111096805997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,512,0.004143110993835661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,1024,0.008972444468074376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,1536,0.01202933324707879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,256,0.002524444419476721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,768,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,1024,0.009317333499590555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,128,0.0022159999029503926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,512,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,768,0.007970666719807519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,512,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,256,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,64,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,32,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,128,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,128,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,6144,0.028524445162879095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,64,0.005881777654091517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,32,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,65536,0.2450559933980306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,51200,0.18830755021837023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,65536,0.14660799503326416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,16384,0.06461955441368951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,65536,0.13463821676042345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,51200,0.11166400379604763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,12288,0.05016622278425428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,51200,0.10267644458346897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,16384,0.03752177622583177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,16384,0.057327999009026416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,10240,0.042583111259672374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,8192,0.03508977757559882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,12288,0.029608001311620075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,8192,0.02365333338578542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,7168,0.030811554855770532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,10240,0.037073777781592474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,8192,0.021872889664438035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,256,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,6144,0.02702666653527154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,8192,0.0314408904976315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,7168,0.019875556230545044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,7168,0.027818666564093694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,6144,0.018010665973027546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,4096,0.02015111181471083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,5120,0.015841777125994366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,5120,0.02178222272131178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,3584,0.01791466606987847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,4096,0.014123555686738757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,4096,0.0199946661790212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,3072,0.015726221932305228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,3584,0.013108444710572561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,3584,0.01719377769364251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,2560,0.013770666387346057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,3072,0.012129777835475074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,3072,0.015199111567603217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,2048,0.011852444046073489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,12288,0.04126222266091241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,2560,0.011077333655622272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,10240,0.02570933269129859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,2560,0.01333688861793942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,1536,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,2048,0.009993777506881291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,1024,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,2048,0.012776888906955719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,1536,0.008631111019187504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,5120,0.023591111103693645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,768,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,512,0.003169777699642711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,1024,0.00814488861295912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,1536,0.01090133355723487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,256,0.0024862223201327855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,6144,0.02390844457679325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,768,0.0074106670088238185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,128,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,512,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,768,0.007298666569921706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,64,0.0019528888579871918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,256,0.006572444405820634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,32,0.0022044444663657085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,128,0.006507555643717448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,256,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,128,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,64,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,65536,0.20423821608225504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,51200,0.16044888231489393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,65536,0.1372071107228597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,16384,0.05718311336305407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,65536,0.12226133876376682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,51200,0.10698311196433173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,12288,0.042578667402267456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,51200,0.09673777553770278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,10240,0.036496000157462224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,16384,0.04822311136457655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,12288,0.027898665931489732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,8192,0.030470222234725952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,12288,0.037456889947255455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,1024,0.008023111356629265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,10240,0.024013333850436743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,7168,0.027317333552572463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,10240,0.031778666708204485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,8192,0.020978666014141504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,6144,0.024284443921513025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,512,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,8192,0.0262071109480328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,5120,0.020759110649426777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,7168,0.019021333919631112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,7168,0.02344977855682373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,6144,0.01713244451416863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,4096,0.017512000269360013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,6144,0.020795555578337777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,5120,0.015623110863897534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,32,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,4096,0.013800889253616333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,3072,0.014076444837782117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,4096,0.016383111476898193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,3584,0.014519999424616495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,3584,0.014557333456145393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,2560,0.012392889294359418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,3072,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,2048,0.009679111341635386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,3072,0.013008889224794177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,2560,0.011743110915025076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,1536,0.00792711145348019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,2560,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,2048,0.009022222624884712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,1024,0.005693333430422678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,2048,0.011200889117187925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,1536,0.007948444121413762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,768,0.0038888889054457345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,1536,0.009512888888518015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,1024,0.008019555773999956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,512,0.002954666606254048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,1024,0.007336888876226213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,768,0.007495111061467065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,256,0.002199999988079071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,768,0.006660444455014334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,512,0.006989333364698622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,3584,0.015560888581805758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,128,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,512,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,256,0.006364444477690591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,64,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,5120,0.018882667024930317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,32,0.0019102222803566193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,256,0.005751111027267244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,128,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,128,0.005437333136796951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,65536,0.16996355851491293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,64,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,32,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,51200,0.12741422653198242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,16384,0.04348799917432997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,16384,0.03593600127432082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,65536,0.07310222254859076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,65536,0.08932177888022529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,12288,0.034323556555642024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,51200,0.05645155244403415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,16384,0.02235377828280131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,12288,0.018912000788582694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,16384,0.034282667769326106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,8192,0.022269333402315777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,12288,0.027497778336207073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,7168,0.01958933307064904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,10240,0.016420443852742512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,10240,0.024337778488794964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,6144,0.01719377769364251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,7168,0.013423999978436364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,7168,0.017480888300471835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,5120,0.014930667148696052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,6144,0.01261333293384976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,6144,0.01681688924630483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,4096,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,5120,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,5120,0.014659555421935188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,3584,0.011008000208271874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,4096,0.013079110946920184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,4096,0.01055466632048289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,3072,0.010008888940016428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,3584,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,51200,0.06903911299175687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,2560,0.008522666990756989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,3072,0.008921778036488427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,3584,0.011739555332395764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,2048,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,3072,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,1536,0.006025777922736273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,2560,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,2048,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,1024,0.004547555413511065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,2560,0.010410666465759277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,2048,0.009383999639087254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,8192,0.015275556180212231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,768,0.003212444484233856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,8192,0.020282665888468426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,1536,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,1024,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,512,0.0025475554996066624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,768,0.0063582224150498705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,256,0.002199999988079071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,768,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,128,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,256,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,256,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,64,0.0018595556418100994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,128,0.005756444401211209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,32,0.0018791111393107309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,128,0.005199111170238919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,65536,0.14732621775733099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,10240,0.026593777868482802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,64,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,32,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,51200,0.11589154932234023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,65536,0.06559999783833821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,1536,0.007107555866241455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,65536,0.08234666453467475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,16384,0.03963911202218797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,51200,0.06236444579230415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,12288,0.031607998741997614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,51200,0.05354488889376322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,16384,0.021664000219768945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,512,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,10240,0.02682755556371477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,16384,0.03343111276626587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,512,0.0057857777509424425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,8192,0.022692445251676772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,12288,0.02677244444688161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,10240,0.015669332610236276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,7168,0.017873777283562552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,10240,0.022614222433831956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,8192,0.014436443646748861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,8192,0.020245333512624104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,7168,0.012984000146389008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,5120,0.013455111119482251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,7168,0.017163554827372234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,6144,0.012061333490742577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,4096,0.011712888876597086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,5120,0.011178666518794166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,5120,0.01425866617096795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,3584,0.010050666828950247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,4096,0.009712888962692684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,4096,0.012712889247470431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,3072,0.009115555220180089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,3584,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,1024,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,2560,0.007912889122962952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,3584,0.01145866678820716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,3072,0.008352888955010308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,3072,0.010336000058386061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,2048,0.006802666518423293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,2560,0.00740444411834081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,1536,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,2560,0.009649777577983009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,2048,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,1024,0.0038204445607132385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,6144,0.015492444237073263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,2048,0.008738666772842407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,1536,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,1536,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,1024,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,512,0.002513777878549364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,1024,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,6144,0.016179554992251925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,768,0.00599644457300504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,256,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,768,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,128,0.0020737778395414352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,512,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,256,0.005852444304360284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,512,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,64,0.0018293333964215384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,256,0.005062222066852781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,128,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,128,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,65536,0.12053244643741184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,64,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,32,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,51200,0.09525510999891494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,65536,0.0737395551469591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,16384,0.03373333480623033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,65536,0.0631493330001831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,12288,0.026782222919993933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,51200,0.05646666553285387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,51200,0.05139288968510098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,10240,0.022913777165942725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,16384,0.020824889341990154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,16384,0.029839999145931665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,8192,0.019659555620617337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,12288,0.024318221542570326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,10240,0.01516266663869222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,7168,0.017509332961506315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,10240,0.020619556307792664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,8192,0.013552889227867126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,6144,0.015489778584904142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,8192,0.01904088921017117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,7168,0.012661333713266583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,5120,0.013742222554153867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,7168,0.016214221715927124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,32,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,6144,0.011760888828171624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,4096,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,6144,0.015134221977657743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,3584,0.009360888765917884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,4096,0.00906311141120063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,5120,0.013400000002649097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,3072,0.008414222134484185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,4096,0.013026666310098437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,12288,0.01761244402991401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,3584,0.008722667064931657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,2560,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,3584,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,2048,0.006292444550328785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,2560,0.0073724447025193115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,768,0.0028515555378463534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,2560,0.008603555460770925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,12288,0.01719288859102461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,1536,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,2048,0.006993778049945831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,1024,0.003359111025929451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,2048,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,1536,0.0069004446268081665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,768,0.0027911110470692315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,1536,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,1024,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,512,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,1024,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,768,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,256,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,5120,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,512,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,128,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,256,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,256,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,64,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,3072,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,128,0.004907555464241239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,32,0.0018800000349680583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,3072,0.009870222045315636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,65536,0.10498399866951837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,64,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,32,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,51200,0.08280977937910292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,65536,0.06696089108784993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,65536,0.060990220970577665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,16384,0.030243555704752605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,51200,0.050880889097849526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,51200,0.04924089047643873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,12288,0.02423200011253357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,16384,0.020236444142129686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,10240,0.02033155494266086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,512,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,16384,0.027074666486846075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,12288,0.016515556308958266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,8192,0.017909333109855652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,12288,0.023009777069091797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,7168,0.015793777174419828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,10240,0.020047111643685233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,8192,0.013187555803192986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,6144,0.0140657772620519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,8192,0.01837688850031959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,7168,0.012139555480745105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,7168,0.015810666812790763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,5120,0.012234666281276278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,6144,0.01129244433508979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,4096,0.010320888625250923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,6144,0.0147733340660731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,5120,0.009716444545321995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,3584,0.008622222476535374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,4096,0.008272888759771982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,5120,0.012820444173283048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,3072,0.0075937774446275495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,4096,0.011911110745535957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,3584,0.008029333419269985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,2560,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,3584,0.01107911103301578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,3072,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,768,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,2560,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,3072,0.008968888885445064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,1536,0.00452888881166776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,2560,0.008245333201355403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,2048,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,1536,0.006575999988449945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,2048,0.007738666401969061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,768,0.002846222163902389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,10240,0.015079110860824585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,1536,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,1024,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,512,0.002467555511328909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,1024,0.006259555617968242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,256,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,128,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,512,0.005768888940413793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,768,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,128,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,512,0.005537777725193236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,64,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,128,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,256,0.005299555758635203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,32,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,128,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,65536,0.09027200274997288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,64,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,32,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,51200,0.06537510951360066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,2048,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,65536,0.059008002281188965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,65536,0.05970755550596449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,16384,0.024726221958796184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,12288,0.019270223047998216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,1024,0.003154666680428717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,51200,0.047545777426825635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,16384,0.01945955554644267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,51200,0.04661422305636936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,10240,0.01681600014368693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,16384,0.027060444156328838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,12288,0.016203555795881484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,8192,0.01479822231663598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,12288,0.021241777473025854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,10240,0.01443555619981554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,7168,0.01292799992693795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,768,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,10240,0.018536888890796237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,8192,0.01260444439119763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,6144,0.011843555503421359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,5120,0.010664888554149203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,8192,0.01687111126052009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,7168,0.011732444167137146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,256,0.005662222289376789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,6144,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,4096,0.009378666679064432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,7168,0.014469333820872836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,3584,0.010595555934641095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,5120,0.009206222163306342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,3072,0.009311999711725447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,4096,0.0081377774477005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,5120,0.01274222218328052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,3584,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,2560,0.008392000363932716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,3072,0.00741244438621733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,3584,0.009429333110650381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,2048,0.007157333195209503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,2560,0.007296889192528195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,3072,0.008298666940795051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,1536,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,2048,0.006812444577614467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,2560,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,1024,0.004416888786686791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,2048,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,1536,0.006667555620272954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,768,0.00350222239891688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,1536,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,512,0.0028888889484935333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,768,0.006055999961164262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,1024,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,256,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,768,0.005850666513045629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,512,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,128,0.0018177777528762817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,512,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,256,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,64,0.0018853334089120228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,256,0.004968889057636261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,32,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,128,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,6144,0.013220444321632385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,128,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,65536,0.0763066675927904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,4096,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,51200,0.05954755677117241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,32,0.00510044437315729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,16384,0.022620444496472675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,65536,0.05714222457673815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,51200,0.04670933220121595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,51200,0.04452622267935011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,12288,0.017511111166742113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,16384,0.019256000717480976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,10240,0.01518133282661438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,1024,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,12288,0.01590844492117564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,8192,0.013080000049538083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,12288,0.020454222957293194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,10240,0.014135999812020196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,7168,0.011952000359694162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,8192,0.012478222449620565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,8192,0.016488000750541687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,6144,0.011086222198274402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,7168,0.011145778000354767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,7168,0.01389955480893453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,5120,0.00978577799267239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,6144,0.009895111123720804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,4096,0.00870577742656072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,64,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,5120,0.008748444418112436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,6144,0.012949333422713809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,3584,0.01015822258260515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,5120,0.011354666617181567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,65536,0.057872891426086426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,3072,0.008604444563388824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,4096,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,2560,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,3584,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,4096,0.01034755590889189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,2048,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,3584,0.008664000365469191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,2560,0.007033777733643849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,1536,0.005124444348944558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,2048,0.0069884442620807225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,2560,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,1024,0.003875555677546395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,2048,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,1536,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,10240,0.01794311073091295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,1536,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,768,0.0035386664999855887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,1024,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,512,0.003002666764789157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,1024,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,768,0.005978666659858491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,256,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,512,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,512,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,128,0.0019724443554878235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,256,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,256,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,64,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,128,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,32,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,128,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,3072,0.00742133292886946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,65536,0.060871111022101514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,3072,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,51200,0.04896622233920627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,32,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,65536,0.056877334912618004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,65536,0.05472266674041748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,16384,0.018862222631772358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,51200,0.04596622122658623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,51200,0.042938666211234204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,16384,0.018922666708628338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,16384,0.023934221929974023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,10240,0.013452444639470843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,768,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,12288,0.018837332725524902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,8192,0.011719110939237805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,10240,0.013960000541475085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,10240,0.016695999436908297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,7168,0.010820444259378644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,8192,0.011734222372372946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,6144,0.009986666341622671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,8192,0.0148062225845125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,7168,0.010574222438865237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,5120,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,16384,0.02458133300145467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,6144,0.009655999640623728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,4096,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,6144,0.012542222109105853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,5120,0.00867555538813273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,3584,0.0075253330998950535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,5120,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,4096,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,4096,0.008949333594905006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,3072,0.009142222503821055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,12288,0.015597333510716757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,2560,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,3584,0.008113777471913232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,3072,0.007361777954631382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,2048,0.005986666513813867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,12288,0.015778667396969266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,3072,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,1536,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,2560,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,2048,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,1024,0.004652444273233414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,2048,0.006752000086837345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,1536,0.006561777657932705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,768,0.003605333467324575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,1536,0.006044444524579578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,1024,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,512,0.0031928889867332247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,1024,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,768,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,64,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,256,0.0025128889828920364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,768,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,128,0.002485333424475458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,512,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,256,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,512,0.0052737775776121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,64,0.0018159999615616268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,32,0.0018159999615616268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,128,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,65536,0.05558844407399496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,3584,0.007895111209816402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,128,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,64,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,51200,0.043248888519075185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,32,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,16384,0.017266665895779926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,65536,0.05582311418321398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,2560,0.0069848886794514125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,65536,0.052220443884531655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,51200,0.04527911212709215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,12288,0.013967111706733704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,10240,0.0120328888297081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,51200,0.04146044452985128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,16384,0.019835554891162448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,12288,0.014965333872371249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,8192,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,10240,0.012708444562223224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,12288,0.017400888933075797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,7168,0.009703111317422655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,10240,0.0151555554734336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,8192,0.011048888994587792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,7168,0.013050666285885705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,6144,0.008904000123341879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,8192,0.013749333719412485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,7168,0.010126222338941362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,7168,0.012455999851226807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,6144,0.009642666412724389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,4096,0.007568000091446771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,256,0.005283555636803309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,5120,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,3584,0.0069226668112807805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,5120,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,4096,0.00831822223133511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,4096,0.007632888853549957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,3072,0.007711999946170383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,3584,0.007689778175618913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,2560,0.007335999773608313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,3584,0.007439999944633908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,3072,0.007501333124107785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,2048,0.005972444597217772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,3072,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,2560,0.007282666862010956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,1536,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,2560,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,2048,0.006903111106819577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,1024,0.0041831110914548235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,2048,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,1536,0.006360000206364526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,768,0.003544888976547453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,1536,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,1024,0.006298666612969504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,512,0.0028986665937635633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,5120,0.008040889269775814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,768,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,256,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,768,0.005292444593376583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,6144,0.010203555226325989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,512,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,128,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,512,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,64,0.0018213333355055915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,128,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,32,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,128,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,65536,0.048060443666246205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,64,0.004967111266321606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,16384,0.01886133352915446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,51200,0.03691644469896952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,65536,0.055088000165091626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,65536,0.05030133326848348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,16384,0.014183110660976835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,51200,0.04451910985840691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,51200,0.04004089037577311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,12288,0.011374221907721626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,1024,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,16384,0.018227555685573153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,16384,0.020917332834667627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,10240,0.010391111175219217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,12288,0.014737778239780001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,8192,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,12288,0.01662133302953508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,10240,0.01279022213485506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,256,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,7168,0.008735111190213097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,10240,0.014369777507252164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,256,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,8192,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,5120,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,7168,0.010074666804737514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,8192,0.012598222328556908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,6144,0.009654222263230218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,7168,0.010715555813577441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,5120,0.008711111214425828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,32,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,3584,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,4096,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,5120,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,3072,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,4096,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,3584,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,2560,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,3072,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,3584,0.007243555453088548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,2048,0.006235555642180973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,3072,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,2560,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,1536,0.005193777796294955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,2560,0.006654222392373615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,1024,0.0038702223036024305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,1536,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,2048,0.006200888918505774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,768,0.003532444437344869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,1536,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,1024,0.006144000010357962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,512,0.0028799999919202593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,1024,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,768,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,256,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,768,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,512,0.005658666706747479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,4096,0.00849955528974533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,512,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,128,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,6144,0.008973333570692275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,256,0.005588444570700328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,64,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,32,0.0019368889431158702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,128,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,65536,0.034450666771994695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,128,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,64,0.005044444567627377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,32,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,65536,0.05476444297366672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,65536,0.04864977796872457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,2048,0.006612444503439798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,16384,0.01202933324707879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,51200,0.04435111085573832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,12288,0.009637333452701569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,16384,0.017720000611411203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,16384,0.017865777015686035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,6144,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,10240,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,12288,0.013405332962671915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,8192,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,12288,0.014763555592960782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,10240,0.011744889120260874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,7168,0.008056888977686564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,8192,0.010670222342014313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,256,0.005103111267089844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,8192,0.010597333312034607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,6144,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,7168,0.00998933364947637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,5120,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,7168,0.009364444348547194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,6144,0.009664889011118147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,4096,0.006015111174848344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,6144,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,51200,0.02732533382044898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,5120,0.008669333325492011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,3584,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,5120,0.008301333420806462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,4096,0.008257778154479133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,3072,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,4096,0.0075626663035816616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,2560,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,3584,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,3072,0.007338666253619724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,2048,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,3072,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,2560,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,1536,0.004609777695602841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,2560,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,2048,0.0069208890199661255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,1024,0.003855111284388436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,2048,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,1536,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,768,0.0032355555643637977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,10240,0.012778666284349231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,1536,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,512,0.002576888849337896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,1024,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,256,0.002181333386235767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,768,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,1024,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,128,0.0021093332519133887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,512,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,768,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,64,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,256,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,512,0.004971555537647671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,32,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,256,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,65536,0.023825777901543513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,128,0.004876444323195351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,51200,0.01958488921324412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,64,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,3584,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,32,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,65536,0.054287999868392944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,16384,0.01312622262371911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,51200,0.03868000043763055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,65536,0.0468408895863427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,51200,0.04366044534577263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,16384,0.015793777174419828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,10240,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,51200,0.03752266698413425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,12288,0.013080000049538083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,8192,0.009194666312800514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,16384,0.015827554795477126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,7168,0.008344000412358178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,12288,0.01314755529165268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,10240,0.011753777662913004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,8192,0.010650667051474253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,10240,0.011039999624093374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,8192,0.009841777384281158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,5120,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,7168,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,6144,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,4096,0.005976000179847081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,5120,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,5120,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,128,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,3584,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,4096,0.007985778152942657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,4096,0.007368889119890001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,3072,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,3584,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,12288,0.0107004443804423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,3584,0.007129777636792924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,2560,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,3072,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,3072,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,2048,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,2560,0.007297777467303806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,1536,0.004553777890072929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,2560,0.006647111227114995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,6144,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,2048,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,1024,0.003850666599141227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,2048,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,7168,0.009977777798970541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,1536,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,768,0.0032035555276605818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,512,0.0028604444944196274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,1536,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,1024,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,256,0.0025048889219760895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,6144,0.008633777499198914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,768,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,1024,0.005679111099905438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,128,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,512,0.00565155554148886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,768,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,256,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,512,0.005157333281305102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,32,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,128,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,256,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,128,0.004866666677925321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,51200,0.015835555063353646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,64,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,32,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,65536,0.05391200052367317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,16384,0.013787555197874704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,65536,0.04571466644605001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,51200,0.04278577698601616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,51200,0.0369777778784434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,16384,0.015792000624868605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,10240,0.008432889150248634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,16384,0.014774221513006421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,12288,0.013074666261672974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,12288,0.01238222254647149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,8192,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,10240,0.011724444727102915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,7168,0.006970666348934174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,10240,0.01106844428512785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,8192,0.01032977799574534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,6144,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,8192,0.009720889230569204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,7168,0.009899555808968013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,5120,0.005624889085690181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,64,0.0018275556051068837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,6144,0.009259555902745988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,4096,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,65536,0.01902222302224901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,6144,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,5120,0.008681777450773451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,3584,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,5120,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,3072,0.005457777943876054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,4096,0.008270222279760573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,2560,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,4096,0.007241778075695038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,3072,0.007323555648326874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,12288,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,2048,0.00423911131090588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,3072,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,1536,0.004507555729813046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,2560,0.006567111031876669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,2048,0.006812444577614467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,1024,0.0035617777870761026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,2048,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,1536,0.006552000012662675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,768,0.0032017777363459268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,1536,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,1024,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,512,0.002562666725781229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,1024,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,768,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,256,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,768,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,7168,0.009308444129096137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,512,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,128,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,512,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,64,0.0018871112002266778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,128,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,256,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,768,32,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,128,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,3584,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,65536,0.02751200066672431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,64,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,3584,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,51200,0.022704000274340313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,32,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,16384,0.00925155563486947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,2560,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,65536,0.04758933186531067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,12288,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,65536,0.04515733321507772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,51200,0.03843199875619676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,10240,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,51200,0.036757333411110774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,12288,0.013120888835854001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,8192,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,12288,0.012286221815480126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,10240,0.012113778127564324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,7168,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,10240,0.010709332923094431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,8192,0.010690666735172272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,6144,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,8192,0.009648888475365108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,7168,0.009761778016885122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,5120,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,7168,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,256,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,6144,0.009324444664849175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,4096,0.0058782220714622075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,6144,0.008631111019187504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,5120,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,3584,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,4096,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,3072,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,4096,0.007384888827800751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,3584,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,2560,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,3072,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,16384,0.015785778562227886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,2048,0.004551110996140374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,16384,0.014479110638300577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,2560,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,1536,0.0035342222286595237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,2048,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,2048,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,1024,0.0028764444092909494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,1536,0.006045333213276333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,1536,0.005956444475385878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,768,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,1024,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,1024,0.005544888890451855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,512,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,768,0.005693333430422678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,5120,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,256,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,768,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,3584,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,512,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,512,0.004872888740566042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,256,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,3072,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,256,0.00463377767139011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,128,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,65536,0.016148444679048326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,128,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,2560,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,64,0.004591111093759537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,32,0.004554666578769684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,65536,0.04616533385382759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,16384,0.0063235556913746735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,65536,0.04235022266705831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,12288,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,51200,0.03730488816897074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,51200,0.03422844409942627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,16384,0.015560888581805758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,10240,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,16384,0.01423377792040507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,12288,0.013051555388503604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,8192,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,10240,0.011729777687125735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,12288,0.012075555821259817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,7168,0.0052266667286554975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,10240,0.010562666588359408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,8192,0.010382222632567087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,6144,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,7168,0.010021333065297868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,8192,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,5120,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,128,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,6144,0.009013333254390294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,4096,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,5120,0.008644444247086843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,512,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,5120,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,4096,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,51200,0.014085332552591959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,3072,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,3584,0.007631111476156447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,3584,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,2560,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,3072,0.007352888584136963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,3072,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,2048,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,2560,0.0070293330483966405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,2560,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,1536,0.003536888708670934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,2048,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,2048,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,1024,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,1536,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,1536,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,7168,0.009132444030708736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,768,0.0025519999778932994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,1024,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,512,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,1024,0.005335111171007156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,3584,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,768,0.005978666659858491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,256,0.0021084443562560612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,768,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,512,0.005619555711746216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,128,0.0018026667336622875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,512,0.004980444494220945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,64,0.001826666709449556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,256,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,128,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,256,32,0.0018542222678661346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,65536,0.01168444421556261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,128,0.0047528888616296975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,64,0.004579555657174853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,51200,0.009707555174827576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,32,0.004679999831649992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,65536,0.042134222057130605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,16384,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,51200,0.0370071099864112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,51200,0.0336079994837443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,12288,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,16384,0.01572533282968733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,16384,0.014502222339312235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,10240,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,6144,0.00830844458606508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,12288,0.013072888884279462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,8192,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,12288,0.011874666644467248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,4096,0.00810933361450831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,7168,0.005231110999981563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,10240,0.011728000309732227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,256,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,8192,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,10240,0.011065777805116443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,6144,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,5120,0.004566222015354368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,7168,0.009673777553770278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,6144,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,7168,0.009325332939624786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,4096,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,5120,0.008656889200210571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,6144,0.00811288919713762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,3584,0.004553777890072929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,5120,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,65536,0.04589244392183092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,4096,0.007991111112965478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,3072,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,2560,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,3584,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,2048,0.0041991112132867174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,3072,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,3584,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,2560,0.007292444507280986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,1536,0.003535110917356279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,3072,0.0068711112770769335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,2560,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,2048,0.0063982220987478895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,1024,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,768,0.002604444407754474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,1536,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,2048,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,512,0.0024986666523747975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,1024,0.005980444451173146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,1536,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,256,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,768,0.005681777579916849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,1024,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,128,0.001840888833006223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,512,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,768,0.005202666752868229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,64,0.0018071112119489247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,256,0.00535111129283905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,512,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,128,32,0.0018648888087934917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,128,0.0053022222386466126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,256,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,65536,0.006806222101052602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,8192,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,51200,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,16384,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,12288,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,64,0.004548444516128964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,32,0.004557333472702238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,10240,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,65536,0.041787554820378624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,16384,0.014116444521480136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,8192,0.004576888763242298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,7168,0.004591111093759537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,12288,0.01204177737236023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,6144,0.004365333252482944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,10240,0.010386666489972008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,5120,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,8192,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,4096,0.0039013334446483185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,7168,0.00907555553648207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,3584,0.004580444345871608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,6144,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,3072,0.0042275554604000514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,5120,0.007828444242477417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,4096,0.0069848886794514125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,2560,0.004541333350870344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,2048,0.0042071110672420925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,3584,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,1536,0.003506666670242945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,3072,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,1024,0.0028080000645584534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,2560,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,2048,0.005903111149867375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,768,0.002512000087234709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,1536,0.005789333333571752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,1024,0.0053431110249625305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,256,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,51200,0.033921778202056885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,128,0.001889777680238088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,768,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,4096,0.007293333609898885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,64,0.0017964444640609955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,32,0.0018026667336622875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,256,0.004569777597983678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,65536,0.00629688882165485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,128,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,51200,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,64,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,16384,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,32,0.004547555413511065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,12288,0.004545777622196409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,10240,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,65536,0.04161955581771003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,8192,0.004187555362780889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,51200,0.03387466735310025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,7168,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,16384,0.014469333820872836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,6144,0.004941333499219683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,12288,0.011750222080283694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,5120,0.003912888881233004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,10240,0.0107022225856781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,4096,0.0038693332009845306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,8192,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,128,0.004862222406599256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,3584,0.0041884444653987885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,3072,0.003891555385457145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,64,512,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,2560,0.00423911131090588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,6144,0.008476444416575963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,5120,0.007672888537247975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,4096,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,3584,0.0070675553547011475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,3072,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,512,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,1024,0.0028133332315418455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,2560,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,768,0.0024746666765875286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,512,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,2048,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,256,0.001871111078394784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,1536,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,128,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,1024,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,64,0.001785777716173066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,512,0.004863111095296012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,32,0.0015093333398302395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,256,0.0045422220395671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,128,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,64,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,32,0.004562666846646203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,7168,0.009356444080670675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,2048,0.004206222378545337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,16384,0.3357493347591824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,7,32,1536,0.0032799999333090256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,51200,0.48610401153564453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,16384,0.18089866638183594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,12288,0.2561928961012098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,51200,1.011074701944987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,12288,0.156150221824646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,10240,0.20212888717651367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,16384,0.13542666700151232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,12288,0.10259289211697048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,8192,0.16531199879116482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,10240,0.1314151154624091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,7168,0.1317840019861857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,8192,0.07032710976070829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,6144,0.11363732814788818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,51200,0.4357306692335341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,8192,0.1180773311191135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,7168,0.09708444277445476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,7168,0.06225066714816623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,10240,0.08627200126647949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,5120,0.09551911221610175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,6144,0.0887119968732198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,4096,0.08032533195283678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,3584,0.06865244441562228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,5120,0.04703288939264086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,6144,0.05639555719163683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,5120,0.0750746660762363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,4096,0.039128889640172325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,4096,0.06500266657935248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,768,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,2560,0.051111999485227794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,3584,0.04383377896414863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,3072,0.0310533344745636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,2048,0.04084622197681003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,3584,0.053575111760033496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,3072,0.049550222025977246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,1536,0.031239112218221027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,2560,0.026766222384240892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,2560,0.04350489046838549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,1024,0.022498667240142822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,2048,0.02327288852797614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,768,0.017619555195172627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,2048,0.03633066680696275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,1536,0.020578665865792167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,1024,0.01757333344883389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,1536,0.02576711111598545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,1024,0.020738666256268818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,768,0.015055111712879606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,256,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,512,0.0127697777416971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,768,0.014538667268223233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,128,0.004975111120276981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,512,0.013083555632167392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,256,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,64,0.004219555606444676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,32,0.0042275554604000514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,128,0.01035199976629681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,256,0.010309333602587381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,3072,0.05961066484451294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,64,0.009026666482289633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,32,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,65536,0.5122755368550619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,512,0.012627555264366997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,51200,0.3886977831522624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,65536,0.9893698162502713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,51200,0.7689368989732531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,16384,0.24866043196784127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,12288,0.18951466348436144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,16384,0.1460906664530436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,128,0.007954667011896769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,12288,0.1259591049618191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,65536,0.4333004421657986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,10240,0.1589591105779012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,10240,0.10485777590009902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,16384,0.11043644613689846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,8192,0.13080978393554688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,51200,0.3602613343132867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,10240,0.07131289111243354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,12288,0.08466044399473403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,8192,0.09246577819188435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,8192,0.057256890667809375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,6144,0.10392355256610447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,7168,0.05069066749678718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,7168,0.07980799674987793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,5120,0.08827910820643108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,6144,0.04363911019431221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,6144,0.07339466942681207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,4096,0.07116622394985624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,3584,0.06307288673188952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,5120,0.06043377849790785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,5120,0.03944622145758735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,4096,0.039239999320771955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,3584,0.029056890143288508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,4096,0.0550782216919793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,3584,0.04473955432573954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,2560,0.04577155576811897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,2048,0.036770668294694685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,3072,0.03161066770553589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,3072,0.0398871103922526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,1536,0.02875022093454997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,2560,0.022878222995334204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,1024,0.020276443825827706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,2560,0.03526666760444641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,2048,0.01958933307064904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,768,0.014896000425020853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,1536,0.017843556072976854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,1536,0.021370665894614324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,512,0.010437332921557956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,1024,0.017824888229370117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,768,0.013599110974205865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,256,0.006256000035338932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,768,0.012757333616415659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,512,0.011086222198274402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,128,0.0031911111954185697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,256,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,512,0.011354666617181567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,256,0.008283555507659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,64,0.00259911103381051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,128,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,3072,0.054543998506334096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,128,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,32,0.0027546667390399506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,64,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,32,0.008195555872387355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,65536,0.32161066267225474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,65536,0.1759751107957628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,2048,0.030172444052166406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,51200,0.24805688858032227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,16384,0.0843226644727919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,7168,0.12188266383277045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,65536,0.15665333800845677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,12288,0.06467466884189181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,51200,0.1300906605190701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,10240,0.05307288964589437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,16384,0.04291022155019972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,16384,0.06488888793521456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,12288,0.03352533446417914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,8192,0.04191999965243869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,12288,0.05138044555981954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,7168,0.0358568893538581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,10240,0.02948444419436985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,10240,0.050789333052105375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,8192,0.02374222212367588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,6144,0.031910220781962075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,7168,0.022089777721299067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,7168,0.03138577938079834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,5120,0.02739022175470988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,6144,0.019720888800091214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,4096,0.022642667094866436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,6144,0.029654221402274236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,5120,0.017885333961910672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,3584,0.019848000672128465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,5120,0.024645333488782246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,1024,0.015404444601800708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,3072,0.01735822194152408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,4096,0.023179555932680767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,3584,0.01891288823551602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,2560,0.014688000082969666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,3584,0.01827822294500139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,3072,0.013139555851618448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,2048,0.012423999607563019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,51200,0.12596266799502903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,2560,0.01236088905069563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,1536,0.009967111051082611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,2048,0.011034666664070554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,2048,0.013795555465751223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,8192,0.037477334340413414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,1024,0.007583111524581909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,1536,0.01000533335738712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,768,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,1536,0.011522666447692446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,512,0.004230222354332606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,1024,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,1024,0.009540444446934594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,768,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,256,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,768,0.006996444529957241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,512,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,128,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,4096,0.015199111567603217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,64,0.0021706666383478376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,512,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,32,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,256,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,128,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,3072,0.017380444539917838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,32,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,64,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,2560,0.015955555770132277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,51200,0.18768355581495497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,65536,0.14627467261420354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,16384,0.064355558819241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,65536,0.12947021590338811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,51200,0.10731022225485908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,12288,0.04986133178075155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,16384,0.054373333851496376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,16384,0.037462221251593694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,10240,0.042504890097512134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,12288,0.029678222205903795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,8192,0.035331557194391884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,12288,0.043367111020618014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,256,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,10240,0.025776000486479864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,7168,0.030761778354644775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,128,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,10240,0.04439111219512092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,6144,0.02700622214211358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,8192,0.022277333670192297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,5120,0.0236124445994695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,8192,0.03167644474241469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,6144,0.018079999420377944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,65536,0.24308445718553331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,4096,0.020350222786267597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,5120,0.016189333465364244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,3584,0.017861333158281114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,5120,0.02071199980047014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,4096,0.014456889695591398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,4096,0.020695999264717102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,51200,0.1117982202106052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,3584,0.013198222551080914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,3072,0.0158906661801868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,2560,0.013787555197874704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,3584,0.01608444419172075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,3072,0.012413333687517377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,2048,0.012106666962305704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,2560,0.011382222175598145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,3072,0.01568711135122511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,1536,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,2560,0.014476444986131458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,2048,0.010136889086829292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,1024,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,1536,0.008954666554927826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,1536,0.01016177816523446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,7168,0.019916445016860962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,1024,0.008216888540320927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,1024,0.00833333366447025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,7168,0.028307557106018066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,512,0.0031475555151700974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,6144,0.025787555509143408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,768,0.007473777565691207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,256,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,512,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,768,0.006595555692911148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,128,0.002177777803606457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,256,0.006335111127959357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,512,0.006353777729802662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,32,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,256,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,128,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,128,0.0054560001525614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,65536,0.20446933640374076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,64,0.0054551110499435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,32,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,65536,0.14377866850958929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,51200,0.16018488672044542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,16384,0.05667733483844333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,65536,0.13025866614447698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,2048,0.012724444270133972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,12288,0.042340444193945996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,51200,0.10653155379825169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,51200,0.0965955588552687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,10240,0.036179555786980525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,768,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,16384,0.035663998789257474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,8192,0.030127998855378892
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,12288,0.02813777658674452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,12288,0.03652000096109178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,7168,0.027070222629441157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,10240,0.02414488957987891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,10240,0.033766223324669734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,8192,0.02068000038464864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,8192,0.026875555515289307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,5120,0.020601777566803824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,7168,0.02296000056796604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,6144,0.017217778497272067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,4096,0.017144888639450073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,6144,0.020844444632530212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,5120,0.015499555402331881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,3584,0.015830222103330824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,5120,0.01773422294192844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,4096,0.013844444519943662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,3072,0.01408088869518704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,4096,0.01630577776167128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,3584,0.015096889601813423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,2560,0.012402666939629449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,3584,0.014465777410401238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,3072,0.011490667031870948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,2048,0.010240889257854885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,3072,0.013450666434235044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,1536,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,2560,0.010572444233629437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,2048,0.009032888544930352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,2560,0.012646222280131446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,1024,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,1536,0.008459555606047312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,16384,0.04774666825930277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,2048,0.011331555744012197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,768,0.0039057777159743835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,1024,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,1536,0.00869511150651508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,6144,0.024007999234729346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,256,0.0022159999029503926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,1024,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,768,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,128,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,512,0.0068959999415609576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,768,0.006575999988449945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,256,0.00655644428398874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,64,0.0018728888697094387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,512,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,32,0.0019120000716712738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,128,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,128,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,65536,0.17122577296362984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,64,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,32,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,65536,0.09055644273757935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,51200,0.1278453270594279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,65536,0.07042844427956475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,51200,0.06883111264970568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,51200,0.055664890342288546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,12288,0.034239110019471906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,16384,0.03481955660714044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,512,0.003056888985964987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,10240,0.026548445224761963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,7168,0.01920355525281694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,12288,0.01833511061138577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,8192,0.02314400010638767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,12288,0.028389334678649902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,10240,0.016523554921150208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,256,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,7168,0.01976711054642995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,8192,0.014473777678277759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,6144,0.017242666747834947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,8192,0.021148444877730474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,7168,0.013404444687896304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,5120,0.015063110325071545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,7168,0.01886311173439026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,6144,0.012527999778588613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,4096,0.012773333324326409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,6144,0.016859556237856548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,16384,0.04351288742489285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,5120,0.011592888997660743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,5120,0.015113777584499784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,3584,0.011002666420406766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,4096,0.010306666294733683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,3072,0.009886222581068674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,4096,0.013454222016864352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,3584,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,2560,0.008657777474986183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,16384,0.02290311124589708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,3072,0.009798222117953831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,3584,0.011946666571829053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,2048,0.007357333269384172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,1536,0.006274666637182236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,3072,0.011069333387745751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,2560,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,1024,0.004550222307443619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,2048,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,2560,0.010317333042621613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,768,0.003206222214632564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,1536,0.007149332927332983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,1024,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,10240,0.025065778030289546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,512,0.0031804444475306403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,1024,0.006771555377377405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,256,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,768,0.00630044440428416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,768,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,512,0.00600888869828648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,128,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,512,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,256,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,64,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,128,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,256,0.005311111195219888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,128,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,65536,0.14623378382788763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,64,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,32,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,51200,0.11576710806952582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,16384,0.03952266772588094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,65536,0.08307733138402303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,65536,0.06566044357087877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,12288,0.0314355558819241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,51200,0.05338399940066867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,16384,0.021519111262427434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,2048,0.009359999663299983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,10240,0.02678933408525255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,12288,0.017671111557218764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,16384,0.03333688775698344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,8192,0.022653333014912073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,12288,0.026434666580624048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,7168,0.01773866679933336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,10240,0.01572444372706943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,10240,0.023824888798925612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,8192,0.013741333451535968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,6144,0.015559110376569959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,8192,0.020251555575264823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,7168,0.012961777547995249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,5120,0.013615111509958902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,32,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,6144,0.01202933324707879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,4096,0.01164177805185318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,5120,0.011036444041464063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,6144,0.016320889194806416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,3584,0.010136889086829292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,5120,0.014324444863531323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,3072,0.009147555463843876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,3584,0.008716444174448649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,4096,0.012827555338541666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,2560,0.00794044468137953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,3584,0.011355555719799466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,3072,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,51200,0.06225688589943779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,3072,0.011023110813564725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,2048,0.006895111252864202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,2560,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,1536,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,1536,0.007017778025733099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,2560,0.009979555176364051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,1024,0.003927999900446998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,2048,0.006986666884687211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,1536,0.006647111227114995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,768,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,1024,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,1536,0.00666133314371109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,512,0.0025351111673646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,1024,0.00654666663871871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,768,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,256,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,512,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,512,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,128,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,256,0.005888888819350137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,4096,0.009720000127951304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,256,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,64,0.0018702221827374564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,32,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,128,0.0057546666098965546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,128,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,65536,0.12098399798075359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,64,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,32,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,51200,0.0949279997083876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,65536,0.07542044586605497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,16384,0.03363377849260966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,65536,0.06403999858432345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,2048,0.00885333369175593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,51200,0.05619821945826212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,51200,0.05156622330347697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,16384,0.021016889148288306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,10240,0.0228426671690411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,16384,0.031209776798884075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,12288,0.01720533271630605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,7168,0.01777688827779558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,8192,0.019727110862731934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,12288,0.025064888927671645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,7168,0.017415111263593037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,10240,0.015229332778188916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,8192,0.013667555318938361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,10240,0.021864000293943617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,6144,0.015471110741297403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,8192,0.019539554913838703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,7168,0.012747555143303342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,5120,0.013733333183659447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,7168,0.01777688827779558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,6144,0.011624000138706632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,4096,0.011008888483047485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,5120,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,6144,0.015595555305480957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,3584,0.009295110901196798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,5120,0.013610666824711693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,4096,0.009128000173303816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,3072,0.008361777497662438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,4096,0.012522666818565793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,2560,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,3072,0.00778311077091429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,12288,0.0266951107316547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,2048,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,2560,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,2560,0.009242666264375051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,1536,0.005304000029961268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,768,0.005635555419656966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,2048,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,1024,0.0034968890249729156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,2048,0.007992000215583378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,1536,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,768,0.0028728888266616394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,1024,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,512,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,1024,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,768,0.006247111078765657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,256,0.0021946666141351065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,768,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,512,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,128,0.0018666666001081467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,512,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,64,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,256,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,128,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,32,0.0018568889548381169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,3584,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,128,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,3584,0.010663111176755695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,65536,0.10531733433405559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,64,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,3072,0.010573333336247338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,51200,0.0827066699663798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,32,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,16384,0.030171556605233088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,65536,0.06712266471650866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,51200,0.04928266670968798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,51200,0.05076444480154249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,16384,0.028076443407270644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,10240,0.020255999432669747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,12288,0.01681511104106903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,12288,0.023178666830062866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,8192,0.017880888448821176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,10240,0.01499466598033905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,10240,0.02090666691462199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,7168,0.015615999698638916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,8192,0.013359111216333179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,8192,0.018378666705555387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,6144,0.013977777626779346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,7168,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,7168,0.015450666348139444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,5120,0.012119111087587146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,6144,0.015133332875039844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,6144,0.011358222199810876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,4096,0.01017155581050449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,65536,0.06095555755827162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,5120,0.00980799976322386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,3584,0.008588444855478074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,16384,0.020282665888468426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,4096,0.008281778130266402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,4096,0.012599111431174807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,3072,0.007668444679843054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,3584,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,2560,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,1536,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,3584,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,2048,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,3072,0.0075866662793689305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,3072,0.009365333451165093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,2560,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,1536,0.004579555657174853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,2048,0.006918222126033571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,2560,0.00868711123863856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,1024,0.003157333367400699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,2048,0.007650666766696506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,768,0.0028204443968004654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,1024,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,1536,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,512,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,1024,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,768,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,256,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,768,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,512,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,128,0.001963555605875121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,512,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,12288,0.024118221468395654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,64,0.0018666666001081467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,5120,0.012770666844315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,32,0.001863111017478837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,256,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,128,0.005703111075692707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,65536,0.08988889058430989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,51200,0.0649813347392612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,32,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,65536,0.058336887094709605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,65536,0.05890578031539917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,16384,0.02464977734618717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,51200,0.047296888298458524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,51200,0.04639644424120585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,16384,0.019595555133289762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,12288,0.01924977699915568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,1536,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,10240,0.016826666063732572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,16384,0.027095999982621934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,8192,0.014829332629839579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,12288,0.016173332929611206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,7168,0.012990222209029727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,10240,0.014390221900410123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,12288,0.022275555464956496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,6144,0.011877333124478659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,10240,0.022848000129063923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,7168,0.011694221860832639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,5120,0.010601777997281818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,256,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,6144,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,7168,0.014671999547216626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,4096,0.009525333013799455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,6144,0.014515555567211576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,5120,0.008972444468074376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,4096,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,5120,0.011735999749766456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,3072,0.00905955582857132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,64,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,3584,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,3584,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,2560,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,3072,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,2048,0.0069120000633928515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,3072,0.007708444363541073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,2560,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,1536,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,2560,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,2048,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,1024,0.0041848888827694785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,2048,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,1536,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,1536,0.006308444258239534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,8192,0.012717333104875354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,768,0.003731555408901639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,1024,0.006292444550328785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,1024,0.0058853332367208265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,512,0.0031386667655573953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,768,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,256,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,512,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,768,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,3584,0.011393778026103973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,128,0.0019537777536445195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,512,0.005464000006516774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,64,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,256,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,32,0.0018568889548381169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,128,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,256,0.00499288903342353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,128,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,65536,0.07614755630493164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,64,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,51200,0.05940177705552843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,32,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,16384,0.022621333599090576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,65536,0.057088891665140785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,51200,0.04661600126160515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,51200,0.0444524453745948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,16384,0.01927911076280806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,12288,0.017605332864655387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,10240,0.015190222197108798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,16384,0.02517066730393304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,12288,0.015847111741701763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,8192,0.017603556315104168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,8192,0.01309866706530253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,10240,0.014150222142537435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,12288,0.01957155598534478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,7168,0.011701333026091257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,10240,0.017871111631393433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,6144,0.011071110765139261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,8192,0.01238933371173011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,7168,0.011376889215575324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,8192,0.016195555528004963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,5120,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,7168,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,6144,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,4096,0.008747555315494537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,6144,0.012660444610648684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,5120,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,3584,0.01035288886891471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,5120,0.01147555559873581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,3072,0.008570666942331526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,4096,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,3584,0.00776088900036282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,2560,0.007479111353556315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,65536,0.05794755617777506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,3584,0.008960000342792934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,2048,0.006396444307433234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,3072,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,3072,0.007729777859316931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,2560,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,1536,0.005164444446563721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,2048,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,2560,0.007330666813585494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,1024,0.0038844446341196695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,2048,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,1536,0.006330666856633292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,768,0.003489777859714296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,1024,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,512,0.0029075555503368378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,768,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,1024,0.005682666682534748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,256,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,768,0.005731555736727185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,512,0.005909333212508096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,128,0.0019680000841617584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,256,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,512,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,4096,0.011402666568756104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,32,0.0018079999006456798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,256,0.00498933345079422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,128,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,65536,0.06121333440144857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,4096,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,64,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,51200,0.04881066746181912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,32,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,16384,0.018900444110234577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,65536,0.05461066630151537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,65536,0.056836446126302086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,12288,0.01578044394652049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,51200,0.0457857780986362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,16384,0.018978666928079393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,51200,0.04275822308328417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,10240,0.013412444127930535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,16384,0.022054221895005967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,12288,0.01571999986966451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,12288,0.01980888843536377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,10240,0.013765333427323235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,1536,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,7168,0.01070666644308302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,10240,0.018596443865034316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,8192,0.011415999796655444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,6144,0.009830222361617619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,8192,0.014766222900814481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,7168,0.010676444404655032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,5120,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,7168,0.013276444541083442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,64,0.0018657777044508194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,6144,0.009483555952707926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,6144,0.012412444584899478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,5120,0.00869422240389718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,128,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,3584,0.007339555356237624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,5120,0.010127999716334874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,3072,0.008376888930797577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,4096,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,4096,0.008786666724416945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,2560,0.007361777954631382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,3584,0.007672000262472365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,3584,0.007699555820888943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,3072,0.007328888608349695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,2048,0.006173333360089197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,2560,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,3072,0.007612444460391998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,2560,0.007256000406212277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,2048,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,1024,0.004176889028814104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,8192,0.011752888560295105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,2048,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,768,0.004222222086456087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,1536,0.005919111271699269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,1024,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,512,0.0031306667046414483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,1024,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,768,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,768,0.005378666851255629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,512,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,4096,0.007704888780911763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,512,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,128,0.0024862223201327855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,256,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,256,0.004955555415815777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,64,0.0018026667336622875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,128,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,32,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,128,0.004975111120276981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,64,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,32,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,1536,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,51200,0.04293333490689596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,65536,0.052526222334967725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,16384,0.01721600029203627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,65536,0.055903997686174184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,1536,0.006599111275540457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,51200,0.045195556349224515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,51200,0.04129688938458761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,16384,0.01870755520131853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,10240,0.012071111136012606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,16384,0.020988444487253826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,12288,0.01482399966981676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,256,0.0025315555847353404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,8192,0.010620444185203977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,12288,0.016866667403115165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,10240,0.013168888787428537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,7168,0.009648000200589498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,10240,0.015838222371207345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,8192,0.011016888750924004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,6144,0.008969777988062965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,7168,0.01036888857682546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,8192,0.013471110827393003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,5120,0.007975999679830339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,4096,0.007676444119877285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,6144,0.009597332941161262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,7168,0.012120000190205045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,65536,0.05501955416467455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,3584,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,6144,0.011026666396194033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,5120,0.008780444661776224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,3072,0.007663999994595845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,4096,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,5120,0.008468444148699442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,3584,0.00775822252035141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,4096,0.007999111380841997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,3072,0.007336888876226213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,2048,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,3072,0.0070186663005087115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,2560,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,1536,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,2560,0.006637333167923822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,2048,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,1024,0.003858666867017746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,12288,0.014031999640994601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,768,0.0035537779331207275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,2048,0.006309333360857434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,1536,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,1024,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,512,0.002955555501911375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,768,0.005889777921968036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,256,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,768,0.005411555369695027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,512,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,512,0.005215999980767568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,128,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,256,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,256,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,64,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,128,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,32,0.0018168888572189545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,128,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,2560,0.006875555548402999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,65536,0.04782933327886793
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,3584,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,64,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,32,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,51200,0.03668533431159125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,16384,0.014336888988812765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,65536,0.055283556381861367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,65536,0.0503404438495636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,12288,0.0114702218108707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,51200,0.044585777653588184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,51200,0.04004888733228048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,16384,0.01830755505296919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,10240,0.01018488903840383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,16384,0.018562666243977018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,12288,0.014072888427310519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,8192,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,10240,0.012267555627557965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,12288,0.017137777474191453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,7168,0.008793777889675563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,10240,0.016754666964213055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,6144,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,8192,0.012038222617573209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,7168,0.010332444475756751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,5120,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,7168,0.01074044406414032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,6144,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,4096,0.008708444734414419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,6144,0.009584888815879822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,5120,0.008669333325492011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,3584,0.008664888640244802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,4096,0.008079999850855933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,4096,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,3072,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,3584,0.00795999997191959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,3584,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,2560,0.0052844443255000645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,1536,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,3072,0.007628444168302748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,2048,0.007272889216740926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,3072,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,2560,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,1024,0.005643555687533484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,1536,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,2560,0.006648889018429651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,2048,0.0069857777820693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,8192,0.011008888483047485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,2048,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,768,0.0035013332962989807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,1536,0.005864000154866113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,1024,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,512,0.002826666666401757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,1024,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,768,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,256,0.002499555548032125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,768,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,128,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,512,0.005653333332803514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,5120,0.00828888929552502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,256,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,64,0.001818666648533609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,512,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,256,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,128,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,65536,0.034254223108291626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,51200,0.027091556125217017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,128,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,64,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,32,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,16384,0.012072888513406118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,65536,0.04836622211668226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,51200,0.04428799947102865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,12288,0.009697777529557547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,16384,0.01700088878472646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,51200,0.03866133424970839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,10240,0.00871288859181934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,16384,0.017824000782436795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,12288,0.01347644461525811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,8192,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,1536,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,7168,0.007920888562997183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,10240,0.011883555187119378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,10240,0.011994666523403592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,8192,0.010788444843557147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,6144,0.007255111303594377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,7168,0.009995555712117089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,8192,0.010053333308961656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,5120,0.006161777923504512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,7168,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,6144,0.0094035557574696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,4096,0.006335999816656113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,5120,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,6144,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,3584,0.005766222046481238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,4096,0.007963555554548899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,4096,0.007495111061467065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,3072,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,3584,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,3584,0.007289778027269576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,2560,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,3072,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,3072,0.0069164443347189165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,2048,0.005631111148330901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,2560,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,1024,0.0041573333243529005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,2560,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,1536,0.0046986668474144405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,12288,0.014652444256676568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,2048,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,1536,0.0063075555695427795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,2048,0.006231110956933763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,768,0.0032133333798911837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,1536,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,1024,0.005975111077229182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,512,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,768,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,5120,0.00813688917292489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,256,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,512,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,512,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,128,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,256,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,65536,0.05464355481995476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,64,0.0018257778137922287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,256,0.004941333499219683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,32,0.0018604445374674266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,128,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,65536,0.02380977736579047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,128,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,64,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,51200,0.01960000064637926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,65536,0.05411200059784783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,1024,0.003636444608370463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,65536,0.046690666013293795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,51200,0.0437520013915168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,51200,0.0373946660094791
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,12288,0.010703999963071613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,16384,0.015582222077581616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,768,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,10240,0.00963466697269016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,12288,0.013186666700575085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,12288,0.012423111332787408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,8192,0.009344889058007134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,10240,0.0118231111102634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,7168,0.008492444124486711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,8192,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,8192,0.010048000348938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,6144,0.007632888853549957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,7168,0.010074666804737514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,7168,0.009349333743254343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,32,0.004960000101062986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,5120,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,6144,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,16384,0.01314222233162986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,6144,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,4096,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,5120,0.008655110994974772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,3584,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,5120,0.008051555189821456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,4096,0.008049777812427944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,3072,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,1024,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,3584,0.00795999997191959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,2560,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,4096,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,16384,0.01601777805222405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,2048,0.005883555445406172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,3584,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,3072,0.007288888924651676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,1536,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,2560,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,3072,0.007000889215204451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,1024,0.0038488888078265717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,2048,0.006581333362393909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,2560,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,1536,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,2048,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,512,0.0028826666788922418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,1024,0.006079999936951532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,1536,0.0059075554211934405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,256,0.0025457777082920074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,1024,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,768,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,512,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,128,0.0021395554973019496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,512,0.005152889009979037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,256,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,256,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,128,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,128,0.004765333400832282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,65536,0.01887999971707662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,64,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,32,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,51200,0.015814221567577787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,65536,0.05410044391949972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,65536,0.04583200150065952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,16384,0.014117333624098035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,51200,0.04357599880960253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,51200,0.036981334288915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,768,0.003338666632771492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,12288,0.009655999640623728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,16384,0.01479644411140018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,10240,0.008290666672918532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,768,0.005632888939645555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,12288,0.01311644415060679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,8192,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,12288,0.012055111428101858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,10240,0.011756444142924415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,7168,0.007002666592597961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,10240,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,32,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,8192,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,6144,0.006341333190600078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,8192,0.009720000127951304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,5120,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,7168,0.00999111102686988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,6144,0.009308444129096137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,4096,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,5120,0.008573333422342936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,5120,0.008063111040327284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,3584,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,4096,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,4096,0.007495111061467065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,3072,0.0052879999081293745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,3584,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,3584,0.007285333342022366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,2560,0.005208000126812193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,3072,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,10240,0.011030221978823343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,3072,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,2048,0.004222222086456087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,2560,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,1536,0.004481777962711123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,2560,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,2048,0.006879111131032308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,1024,0.0035439998739295532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,2048,0.0063031112982167145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,1536,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,1536,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,1024,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,7168,0.009297777381208207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,1024,0.005555555638339784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,768,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,256,0.002151111140847206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,768,0.005399999933110342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,512,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,128,0.0020151111400789684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,512,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,256,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,64,0.0018115554832749898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,256,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,128,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,16384,0.015839111473825242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,65536,0.026546667019526165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,64,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,51200,0.02188266648186578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,32,0.004592888885074192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,65536,0.05084266596370273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,16384,0.009434666898515489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,65536,0.04545244574546814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,51200,0.03867111272282071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,12288,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,51200,0.03660888804329766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,768,0.003051555612021022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,10240,0.006641777853171031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,16384,0.015828443898095023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,16384,0.01443288889196184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,768,512,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,8192,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,10240,0.012072888513406118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,12288,0.012059555285506778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,7168,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,10240,0.010969777901967367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,8192,0.010688888529936472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,6144,0.005975111077229182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,7168,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,5120,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,7168,0.009304888546466827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,128,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,6144,0.008678221868144141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,4096,0.005648889061477449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,5120,0.008618666893906063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,3584,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,5120,0.008299555215570662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,4096,0.00792711145348019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,3072,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,3584,0.007967111137178209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,3584,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,3072,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,2560,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,12288,0.013080000049538083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,6144,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,2048,0.004816888935036129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,2560,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,8192,0.009725333087974125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,1536,0.0035964445107513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,6144,0.009073778159088558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,2048,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,1024,0.0028719999310043124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,1536,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,1536,0.005752888818581899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,768,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,1024,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,1024,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,4096,0.007294221884674496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,512,0.0025555555605226094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,768,0.005631999837027655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,256,0.0022942223068740633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,768,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,512,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,128,0.0019199999256266488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,3072,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,512,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,64,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,256,0.004631999880075455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,128,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,512,32,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,128,0.004534222185611725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,65536,0.016876444220542908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,2560,0.006261333409282897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,51200,0.014464888307783337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,64,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,2048,0.006721777634488211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,32,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,16384,0.006587555425034628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,51200,0.037300444311565824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,65536,0.042091555065578885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,12288,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,51200,0.03429066803720262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,16384,0.015629332926538255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,16384,0.014326221413082547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,12288,0.0119991112086508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,8192,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,10240,0.011706666813956367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,10240,0.011039111349317761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,7168,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,8192,0.010454222559928894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,8192,0.00960355583164427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,256,0.0052968888647026485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,6144,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,7168,0.010006222460005019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,5120,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,6144,0.009317333499590555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,4096,0.004252444538805219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,5120,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,6144,0.008345777789751688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,3584,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,5120,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,65536,0.04593511091338264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,4096,0.008263111114501953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,3072,0.004639111045334074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,3584,0.007678222325113084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,4096,0.007028444773621029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,2560,0.004546666724814309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,10240,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,3072,0.007311999797821045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,2048,0.004521777646409141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,12288,0.013055110971132914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,2560,0.00692622239391009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,1536,0.0035004446076022256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,2560,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,1024,0.002875555513633622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,1536,0.0064488889442549805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,2048,0.005965333431959152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,768,0.002534222271707323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,1536,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,1024,0.00600622221827507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,512,0.002469333302643564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,1024,0.0052764444715446895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,768,0.005623111294375525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,256,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,768,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,512,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,128,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,512,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,256,0.005312888986534542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,64,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,256,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,256,32,0.0017946666727463405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,128,0.005373333477311664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,65536,0.011446221835083432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,128,0.004606222112973531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,64,0.004607111215591431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,3584,0.006852444261312485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,51200,0.009702222214804756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,32,0.004584889031118817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,65536,0.04585955540339152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,16384,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,3072,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,65536,0.04179199867778354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,12288,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,51200,0.03700088792377048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,2048,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,10240,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,16384,0.015830222103330824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,12288,0.01278222186697854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,16384,0.014137778017255994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,8192,0.005046222358942032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,10240,0.011791999969217511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,7168,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,10240,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,8192,0.010407999985747868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,7168,0.009333333207501305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,6144,0.004562666846646203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,8192,0.009666666388511658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,7168,0.009701333112186855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,5120,0.0042133331298828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,6144,0.009315555294354757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,4096,0.004261333495378494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,5120,0.008420444197124904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,6144,0.008959111240175035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,3584,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,5120,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,4096,0.00795377790927887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,3072,0.004261333495378494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,4096,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,3584,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,2560,0.004548444516128964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,3584,0.007318221860461765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,3072,0.007256888680987888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,2048,0.0042133331298828125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,3072,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,2560,0.007278222176763747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,51200,0.033935109774271645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,1536,0.0035111109415690103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,2560,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,1024,0.002809777855873108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,2048,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,12288,0.01202044470442666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,2048,0.005976888868543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,1536,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,768,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,512,0.002293333411216736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,1024,0.005969777703285217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,768,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,1024,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,256,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,512,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,768,0.004972444640265571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,7168,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,512,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,256,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,64,0.0017626666360431248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,32,0.001782222133543756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,256,0.0047217777205838096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,128,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,65536,0.006582222051090664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,51200,0.006227555374304454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,128,0.004604444321658877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,16384,0.005296000176005893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,64,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,12288,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,32,0.004868444469239977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,10240,0.004552888787455029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,65536,0.04179111123085022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,8192,0.005020444591840108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,51200,0.033526221911112465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,7168,0.004573333180612988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,16384,0.014159111513031853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,6144,0.004285333471165763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,12288,0.012059555285506778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,5120,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,8192,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,4096,0.0042364444169733255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,7168,0.008963555925422246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,1536,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,3072,0.0042035554846127825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,5120,0.007965332931942409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,2560,0.004527111020353106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,128,128,0.001810666587617662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,4096,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,2048,0.0041991112132867174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,1536,0.00350222239891688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,3584,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,1024,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,3072,0.006575999988449945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,768,0.002492444382773505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,2560,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,512,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,2048,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,256,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,1536,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,128,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,1024,0.0053119998839166425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,64,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,768,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,32,0.001806222316291597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,512,0.005061333378156026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,256,0.005245333330498801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,65536,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,128,0.0046977777447965415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,64,0.0046311111913787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,10240,0.010992888775136737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,16384,0.005534222142563925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,32,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,12288,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,6144,0.008512000242869059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,65536,0.04177422324816386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,10240,0.004629333400064045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,51200,0.03393955694304572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,8192,0.004239999999602635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,7168,0.004552888787455029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,16384,0.014169777433077494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,6144,0.004592888885074192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,12288,0.012044444680213928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,5120,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,10240,0.011040888726711273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,4096,0.0042160000238153665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,8192,0.009672000176376766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,7168,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,3584,0.004229333251714706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,3072,0.004249777644872665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,5120,0.00795199970404307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,2560,0.004543111142185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,4096,0.0073928890956772705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,2048,0.004575110971927643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,3584,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,3072,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,1536,0.0032008888406885993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,1024,0.0027804445061418745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,2560,0.006322666588756774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,768,0.002475555572244856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,51200,0.006263999889294307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,2048,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,512,0.002199999988079071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,256,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,1536,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,1024,0.0053342220683892565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,64,3584,0.00452888881166776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,64,0.0017946666727463405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,768,0.005281777845488654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,32,0.001793777777089013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,256,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,128,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,64,0.004540444248252445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,32,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,6144,0.008611555728647444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,16384,0.3352026674482558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,16384,0.21595022413465711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,51200,0.4867288801405165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,12288,0.25556710031297475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,16384,0.13510933187272814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,6,32,128,0.0018168888572189545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,51200,1.0771866904364693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,10240,0.21420978175269234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,512,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,12288,0.10246133142047459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,8192,0.1641777753829956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,10240,0.08778844277064006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,7168,0.14254932933383518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,8192,0.06888533300823636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,51200,0.444668451944987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,8192,0.11800532870822483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,6144,0.11356443829006618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,7168,0.06304177973005506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,5120,0.09534577528635661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,6144,0.08790666527218288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,7168,0.10659999979866876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,5120,0.04684622089068095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,4096,0.08063022295633952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,5120,0.07667555411656697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,3584,0.06875644127527873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,4096,0.06536266538831922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,4096,0.03901333279079861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,12288,0.17767555183834502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,3584,0.03492533498340183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,3584,0.05199999941719902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,2560,0.050760888391070895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,10240,0.12768977218204075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,3072,0.050591998630099826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,2048,0.04101244608561198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,1536,0.03228444523281521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,2560,0.0273324449857076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,2048,0.022629333866967097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,1024,0.02239377796649933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,2048,0.03839733203252157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,1536,0.022608000371191237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,2560,0.03854044609599643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,768,0.01754844519827101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,1024,0.018227555685573153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,1536,0.027662222584088642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,512,0.01273155543539259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,1024,0.020787555310461257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,768,0.015450666348139444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,256,0.0075191110372543335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,768,0.01718488832314809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,128,0.005079999979999331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,512,0.01291822228166792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,256,0.010370666782061258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,512,0.013298666311634911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,64,0.0042453333735466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,3072,0.059156444337632924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,128,0.01036088913679123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,256,0.009992889232105678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,32,0.004228444563017951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,128,0.008302222523424361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,64,0.00904266701804267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,3072,0.03104355600145128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,32,0.00831200016869439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,65536,0.5133431222703722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,51200,0.38861245579189724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,6144,0.05424355467160543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,51200,0.7976560062832303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,65536,1.0093893475002713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,16384,0.2570942242940267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,16384,0.15028711160024008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,51200,0.3161982165442573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,16384,0.11176889472537571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,12288,0.1218951145807902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,12288,0.18972444534301758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,12288,0.0836942195892334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,10240,0.15786666340298122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,65536,0.4336000018649631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,8192,0.12968356079525417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,7168,0.11385067303975423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,6144,0.10449955860773723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,10240,0.1019235584470961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,10240,0.07095022333992852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,8192,0.0582631097899543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,5120,0.088736891746521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,7168,0.08408711353937785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,7168,0.05327555537223816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,6144,0.0714284446504381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,4096,0.07138933075798883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,6144,0.04448799954520332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,5120,0.03865155577659607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,3584,0.06272177563773261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,3072,0.05449155635303921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,4096,0.04338399900330437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,4096,0.05511733227305942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,3584,0.03298755486806234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,2560,0.04563644528388977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,3584,0.0427706672085656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,3072,0.026128000683254663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,3072,0.04279999931653341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,1536,0.028622223271263972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,2048,0.01956622302532196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,2048,0.030015110969543457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,1024,0.020003555549515616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,2560,0.03256266646915012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,1536,0.023611555496851604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,1536,0.019003555178642273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,768,0.014877332581414117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,1024,0.01538222200340695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,512,0.0103493332862854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,8192,0.09346310959921943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,1024,0.01759111053413815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,768,0.01367911116944419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,256,0.006256888724035687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,768,0.014477334088749357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,128,0.003212444484233856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,512,0.010967111421955956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,256,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,512,0.011624889241324531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,64,0.002521777732504739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,256,0.00831733312871721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,32,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,128,0.009382222261693742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,64,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,2048,0.03641155693266127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,32,0.007306666837798224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,65536,0.3185466660393609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,2560,0.022965333527988855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,16384,0.08460888597700332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,65536,0.17425421873728433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,51200,0.24755999777052137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,12288,0.06480711036258273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,51200,0.13011822435590956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,16384,0.041871110598246254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,65536,0.15817422337002224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,16384,0.06424355506896973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,10240,0.05328177743487888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,51200,0.11448710494571263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,8192,0.041859553919898145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,12288,0.033193777004877724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,10240,0.029639999071757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,7168,0.036502222220102944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,10240,0.043004443248113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,12288,0.050457778904173106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,5120,0.0678995582792494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,6144,0.03154044349988302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,8192,0.03696711195839776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,5120,0.027110222313139174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,7168,0.021886222892337378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,8192,0.024506666594081458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,4096,0.022631999519136217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,6144,0.01957155598534478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,5120,0.01868355605337355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,6144,0.028631998433007136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,3584,0.019575110740131803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,128,0.007341333561473423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,5120,0.02493422230084737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,4096,0.015399111641777886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,3072,0.017286222842004564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,2560,0.01482399966981676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,4096,0.022683555881182354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,3584,0.014140443669425117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,3072,0.01313066648112403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,2048,0.01239911135700014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,3584,0.018072888255119324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,2560,0.012215110990736218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,1536,0.009641778137948778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,2560,0.014725334114498563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,1024,0.007275555696752336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,1536,0.010804444551467896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,2048,0.014117333624098035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,768,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,1536,0.012063110868136087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,1024,0.009259555902745988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,1024,0.009694221946928237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,768,0.008064000142945183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,768,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,256,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,512,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,512,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,7168,0.03250222073660956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,128,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,256,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,64,0.0019164443429973389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,256,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,128,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,128,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,64,0.005949333310127258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,3072,0.017952000101407368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,32,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,2048,0.011037333144081963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,65536,0.2394177648756239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,51200,0.18705422348446318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,65536,0.14678755071428087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,512,0.0041955556306574075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,65536,0.1298560036553277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,51200,0.11185956001281738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,12288,0.049823111957973905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,16384,0.0374257763226827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,10240,0.04243555665016174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,12288,0.029842668109469946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,16384,0.054021331999037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,51200,0.10210044516457452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,12288,0.044304887453715004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,8192,0.03529333406024509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,7168,0.030615111192067463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,32,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,10240,0.025962667332755193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,10240,0.03645066751374139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,8192,0.02182577715979682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,8192,0.03279733326699998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,7168,0.028370665179358587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,5120,0.023584889041052923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,4096,0.020311110549502902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,6144,0.018056000272432964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,6144,0.024544888072543677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,5120,0.01664444473054674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,3584,0.017682666579882305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,5120,0.023152000374264185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,4096,0.020167110694779288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,3072,0.015785778562227886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,3584,0.01475911173555586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,16384,0.06446488698323567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,3584,0.016195555528004963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,3072,0.012262221839692859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,2048,0.012073777616024017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,3072,0.016494222813182406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,2560,0.011223999990357293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,1536,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,2048,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,2560,0.013632888595263163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,1024,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,2048,0.012864888542228274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,1536,0.009302222066455418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,1536,0.01090311093462838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,768,0.005128889034191768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,6144,0.026906667484177485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,7168,0.01956355571746826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,512,0.0031617778456873367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,1024,0.007978666987684038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,768,0.007491555478837755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,256,0.002514666567246119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,768,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,128,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,512,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,256,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,512,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,64,0.0019617778145604664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,4096,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,256,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,128,0.0064088888466358185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,128,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,2560,0.013554666605260638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,64,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,65536,0.2036879989835951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,32,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,51200,0.15996000501844618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,65536,0.1401324404610528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,16384,0.056361774603525795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,51200,0.0986657804912991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,51200,0.10722400082482232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,12288,0.042136000262366406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,16384,0.04599644409285652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,1024,0.00830577810605367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,10240,0.03618311219745212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,12288,0.028622223271263972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,12288,0.03695644603835212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,8192,0.029920889271630183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,10240,0.0240666667620341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,10240,0.03110399842262268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,7168,0.02711822258101569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,32,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,8192,0.02124888863828447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,6144,0.024004444479942322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,8192,0.027620444695154827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,7168,0.01920000049802992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,5120,0.02051999999417199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,6144,0.017162667380438913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,7168,0.024289778537220422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,6144,0.02088000045882331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,5120,0.017424000634087455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,3584,0.015451555450757345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,5120,0.018223999275101554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,4096,0.01424711114830441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,3072,0.013776889277829064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,4096,0.016470222009552848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,3584,0.012816888590653738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,3584,0.01441511180665758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,2560,0.012447111308574677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,16384,0.035164445638656616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,3072,0.011878222227096558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,2048,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,3072,0.014069333672523499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,2560,0.010508444574144153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,1536,0.007925333248244392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,2048,0.00902133352226681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,2560,0.0120666664507654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,1024,0.005694222119119432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,2048,0.011037333144081963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,1536,0.008967999782827165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,1024,0.007920888562997183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,1024,0.0075902218619982404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,512,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,768,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,768,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,256,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,4096,0.017502221796247695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,512,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,128,0.00216977774269051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,512,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,256,0.005745777653323279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,64,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,128,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,32,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,65536,0.12550844086541071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,128,0.0052533331844541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,65536,0.17121599780188668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,64,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,32,0.005320000151793162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,51200,0.12642133235931396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,65536,0.07014577918582492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,65536,0.09193155500623916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,16384,0.04318488968743218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,12288,0.033921778202056885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,1536,0.009345778160625035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,51200,0.055504000849194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,16384,0.023077333966890972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,10240,0.027059555053710938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,16384,0.03527999917666117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,12288,0.018295110927687753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,8192,0.02274666726589203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,10240,0.016388444436921012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,7168,0.01959911154376136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,10240,0.02459377712673611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,256,0.0063999998900625445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,6144,0.0170817772547404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,8192,0.020599111914634705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,7168,0.014004444082578024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,5120,0.014769777655601501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,7168,0.018911111685964797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,6144,0.012545777691735161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,4096,0.01277599980433782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,6144,0.016511110795868766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,5120,0.012778666284349231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,3584,0.010991999672518836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,5120,0.014422222971916199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,4096,0.010495999621020423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,3072,0.009957333405812582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,4096,0.013329777452680798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,768,0.003842666745185852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,3584,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,2560,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,3584,0.011691555380821228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,51200,0.06864177518420748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,2048,0.007371555599901412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,3072,0.01125955581665039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,2560,0.008316444853941599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,1536,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,2560,0.010048000348938836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,2048,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,2048,0.009088888764381409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,1024,0.004573333180612988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,12288,0.030065778228971694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,1536,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,768,0.003109333415826162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,1536,0.007885333564546373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,1024,0.006643555644485686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,512,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,768,0.006335111127959357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,1024,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,768,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,512,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,128,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,512,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,256,0.00590844452381134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,64,0.0018657777044508194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,128,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,32,0.0019706665641731685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,128,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,65536,0.14671022362179226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,3072,0.00868266655339135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,64,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,32,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,51200,0.11509511205885146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,65536,0.06595377789603339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,65536,0.08363199896282619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,51200,0.06231466929117838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,16384,0.03935022155443827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,51200,0.05355111095640394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,12288,0.03124444352255927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,256,0.0022560000005695554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,16384,0.021878222624460857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,10240,0.026512000295850966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,16384,0.03268622358640035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,12288,0.017885333961910672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,8192,0.022650667362742957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,12288,0.030421333180533514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,7168,0.017651556266678702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,10240,0.015569777952300178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,256,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,10240,0.022322666313913133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,8192,0.014112888111008538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,6144,0.015632000234391954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,8192,0.019559999306996662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,7168,0.013053333593739403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,6144,0.01205599970287747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,7168,0.018214222457673814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,4096,0.0120328888297081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,5120,0.011014222270912595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,6144,0.016128000285890367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,3584,0.010088000032636855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,5120,0.01407999959256914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,4096,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,3072,0.009032888544930352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,3584,0.008565333154466417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,2560,0.00795288880666097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,3072,0.008429333567619324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,8192,0.014722666806644864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,3584,0.011143999795118967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,2048,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,1536,0.005798222290145026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,3072,0.011040888726711273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,2560,0.007626666790909237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,1024,0.0038684445122877755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,2048,0.0069644442862934535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,1536,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,768,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,1536,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,1024,0.0063235556913746735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,512,0.002581333327624533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,1024,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,768,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,256,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,768,0.006004444426960415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,5120,0.013423999978436364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,512,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,128,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,256,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,64,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,128,0.005886222339338726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,32,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,128,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,4096,0.012841777669058906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,65536,0.1208160056008233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,32,0.005219555563396878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,51200,0.09502133395936753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,2560,0.010323555933104621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,65536,0.07464977767732409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,2048,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,51200,0.05152622196409437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,16384,0.030559112628300984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,16384,0.020769778225156996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,12288,0.026770666241645813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,512,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,256,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,12288,0.017305778132544625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,10240,0.02275733318593767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,12288,0.024520888924598694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,8192,0.01963911122745938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,64,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,10240,0.020948444803555805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,7168,0.017335999343130324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,8192,0.013444444371594323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,8192,0.018995554910765756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,6144,0.015541333291265698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,7168,0.012704888979593912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,16384,0.03362933463520474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,7168,0.016679111454221938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,5120,0.013766222529941134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,6144,0.011663999822404651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,51200,0.05607910950978597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,6144,0.0151591118839052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,5120,0.010397333237859938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,3584,0.009332444104883406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,5120,0.013408000270525614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,4096,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,3072,0.00831288927131229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,4096,0.012732444538010491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,3584,0.008569777839713627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,2560,0.007359999749395583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,3584,0.010750222537252637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,3072,0.007745777567227681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,2048,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,3072,0.010766222245163388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,2560,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,1536,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,10240,0.015611555841233997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,2560,0.008835555778609382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,1024,0.0034026667061779234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,2048,0.00702133360836241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,2048,0.008180444439252218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,1536,0.006811555474996567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,768,0.002874666617976295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,1536,0.007309333317809635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,512,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,1024,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,65536,0.06333688894907634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,768,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,256,0.0022053333620230355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,768,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,512,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,4096,0.010934222075674268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,512,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,256,0.005639111002286275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,64,0.001855111163523462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,32,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,256,0.0052266667286554975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,128,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,128,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,65536,0.1042319999800788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,64,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,51200,0.08251200119654338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,32,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,16384,0.029920889271630183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,65536,0.06611555814743042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,65536,0.06100710895326403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,12288,0.023992889457278784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,51200,0.049248887432946094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,51200,0.050739556550979614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,10240,0.020215999748971727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,16384,0.028583112690183852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,16384,0.020279111133681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,12288,0.016513778103722464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,8192,0.017906667457686532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,7168,0.01588888963063558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,12288,0.0237137774626414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,1024,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,10240,0.014767110347747803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,6144,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,8192,0.013060444758998023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,5120,0.012289777398109436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,10240,0.019910222954220243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,7168,0.012301333248615265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,8192,0.017849778135617573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,4096,0.009948444035318162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,6144,0.011094222466150919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,128,0.0021475555582178966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,3584,0.008640000389681922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,6144,0.01442577772670322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,5120,0.00982488857375251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,3072,0.007667555577225155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,4096,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,2560,0.0069306666652361555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,5120,0.012720888687504662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,3584,0.008024888734022776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,4096,0.012044444680213928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,2048,0.005968889014588461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,3072,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,3584,0.010337777435779572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,1536,0.00452533322903845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,2560,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,3072,0.010009778042634329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,1024,0.0031617778456873367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,2048,0.0070222218831380205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,2560,0.008272888759771982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,1536,0.006521777974234686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,2048,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,1024,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,1536,0.006912888752089606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,768,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,768,0.005897777775923411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,256,0.0021342221233579847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,512,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,512,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,128,0.001965333397189776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,256,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,256,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,64,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,7168,0.016528000434239704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,128,0.0048942222363419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,64,0.005047111047638787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,32,0.005195555587609609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,65536,0.0892088876830207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,768,0.0028248888750871024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,512,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,51200,0.06480889187918769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,65536,0.058851553334130176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,16384,0.024615110622511968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,51200,0.04753155509630839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,51200,0.04632622334692213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,12288,0.019242667489581637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,16384,0.027120888233184814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,16384,0.019555555449591745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,10240,0.016597333881590102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,128,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,12288,0.01606666710641649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,32,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,8192,0.01479288935661316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,10240,0.014434667097197639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,10240,0.018550222118695576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,7168,0.013027555412716336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,8192,0.012737777498033313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,6144,0.011756444142924415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,8192,0.01761777698993683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,7168,0.011428444749779172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,5120,0.010640888578361936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,6144,0.010390222072601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,7168,0.0148026661740409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,4096,0.009416888985368941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,6144,0.013930666777822705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,3584,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,5120,0.009008888569143083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,4096,0.00814488861295912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,5120,0.012063999970753988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,3072,0.00962222201956643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,4096,0.011055111057228513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,3584,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,2560,0.008293333152929941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,3584,0.008983111215962304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,2048,0.00721244431204266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,3072,0.007634667058785756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,3072,0.00869511150651508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,1536,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,2560,0.008016000191370646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,2048,0.006913777854707506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,1024,0.0042444442709287005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,12288,0.02163644466135237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,2048,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,768,0.003863999827040566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,1536,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,1024,0.006313777632183499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,512,0.0029537777105967202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,1024,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,768,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,768,0.005703111075692707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,256,0.0022035555707083806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,512,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,128,0.0018604445374674266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,512,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,256,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,1024,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,256,0.005251555393139522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,128,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,32,0.0018524444765514797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,128,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,65536,0.07818310790591769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,64,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,51200,0.05946133534113566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,32,0.005240889059172736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,2560,0.0070293330483966405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,16384,0.02256800068749322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,65536,0.05781244569354587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,65536,0.056927998860677086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,12288,0.01748533381356133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,51200,0.04654666781425476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,1536,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,51200,0.04456355505519443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,10240,0.015061333775520325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,16384,0.025051555699772302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,12288,0.01587377819750044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,10240,0.01440355512830946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,12288,0.022978666755888198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,10240,0.017883555756674874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,7168,0.01182577759027481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,8192,0.012424888710180918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,8192,0.016512889001104567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,65536,0.059976001580556236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,7168,0.011354666617181567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,7168,0.014762666490342883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,5120,0.009672889278994666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,6144,0.009744889206356471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,6144,0.012048000262843238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,4096,0.008645333349704742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,5120,0.009033777647548253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,5120,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,3584,0.009879111415810054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,4096,0.008360000120268928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,4096,0.009016888837019602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,16384,0.01920622256067064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,3584,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,8192,0.013162666724787818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,3584,0.008228444390826756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,2560,0.00830666638082928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,3072,0.007644444704055786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,3072,0.008573333422342936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,2048,0.006858666737874349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,2560,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,1536,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,2560,0.007610666255156199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,64,0.0018595556418100994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,1024,0.0041955556306574075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,2048,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,1536,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,768,0.003478222423129611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,1536,0.0063386667105886675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,1024,0.006200888918505774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,512,0.00281866660548581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,1024,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,768,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,256,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,768,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,512,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,512,0.005501333210203383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,256,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,64,0.0018088887963030075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,3072,0.008978666530715095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,256,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,32,0.0018008889423476325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,128,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,128,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,65536,0.06111911270353529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,51200,0.04864177770084805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,64,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,32,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,65536,0.05692444245020548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,65536,0.05458755625618828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,6144,0.011075555450386472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,51200,0.04587199952867296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,12288,0.015453333655993143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,51200,0.042690667841169566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,16384,0.019104000594880845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,10240,0.013474666410022311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,16384,0.02235288918018341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,12288,0.015646222564909194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,12288,0.019547555181715224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,10240,0.014131555954615274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,10240,0.016488000750541687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,7168,0.010851555400424533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,8192,0.01145155562294854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,128,0.0019057778020699818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,8192,0.014109333356221518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,7168,0.011221333510345884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,5120,0.009111999637550777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,6144,0.009692444569534725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,6144,0.011770666473441653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,4096,0.00796977761718962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,5120,0.008777778181764815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,5120,0.010552888943089379
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,3584,0.0074346669846110875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,4096,0.00830844458606508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,16384,0.01888799998495314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,3072,0.008607111043400234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,4096,0.009277332988050249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,2048,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,3584,0.007993777592976889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,2560,0.007449777589903937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,3584,0.007671111159854465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,2048,0.006211555666393704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,3072,0.007618666523032718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,3072,0.007674666742483775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,2560,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,8192,0.011707555916574268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,1536,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,1024,0.004754666652944353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,2560,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,2048,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,768,0.003867555409669876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,1536,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,2048,0.0063973334100511335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,512,0.0031582222630580268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,1024,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,1536,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,6144,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,256,0.0028088889602157804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,768,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,512,0.005943999936183293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,768,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,7168,0.01349511080318027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,128,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,256,0.005578666511509154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,64,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,512,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,32,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,256,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,128,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,128,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,64,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,32,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,65536,0.05580621957778931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,16384,0.01721600029203627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,65536,0.05199644300672743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,51200,0.045178665055168994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,51200,0.041287110911475286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,16384,0.02144444485505422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,12288,0.015160888433456421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,10240,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,1024,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,10240,0.013071999781661563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,8192,0.010435555544164447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,10240,0.015107555521859063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,65536,0.05552088883188036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,7168,0.009679999616410997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,8192,0.011036444041464063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,51200,0.04268177681499057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,8192,0.013780444032616086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,6144,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,7168,0.010310221877362993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,5120,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,7168,0.012131555212868584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,6144,0.009476444787449306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,12288,0.013621333572599622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,6144,0.010738666686746808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,3584,0.007012444237867991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,5120,0.009321778184837764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,4096,0.008250666989220513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,3072,0.007589333587222629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,4096,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,12288,0.01722666621208191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,3584,0.007684444387753804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,2560,0.0068195557428730865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,3584,0.0074720001882976955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,2048,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,3072,0.007268444531493717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,2560,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,1536,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,2560,0.006636444479227066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,2048,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,1024,0.00423555572827657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,1536,0.006308444258239534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,1536,0.006037333359320958
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,768,0.003567111161020067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,1024,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,1024,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,512,0.0028479999552170434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,768,0.005884444548024072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,4096,0.007648888561460707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,16384,0.018800000349680584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,256,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,5120,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,128,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,512,0.005628444254398346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,256,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,512,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,64,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,32,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,128,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,256,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,65536,0.04695999953481886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,128,0.004953777624501122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,51200,0.03644977675543891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,64,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,3072,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,32,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,65536,0.055231998364130654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,65536,0.049590223365359835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,12288,0.011593778100278644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,51200,0.044659554958343506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,51200,0.040064887868033514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,10240,0.01035733355416192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,16384,0.020612445142534044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,12288,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,8192,0.00941422250535753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,12288,0.016565332810084026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,10240,0.012253333297040729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,7168,0.008976889153321585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,10240,0.013438222308953604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,8192,0.011024889018800525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,768,0.005459555735190709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,6144,0.007920888562997183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,8192,0.01278222186697854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,5120,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,7168,0.010112000008424124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,7168,0.010608000059922537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,6144,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,4096,0.008794666992293464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,5120,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,5120,0.00869599978129069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,3584,0.008283555507659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,4096,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,16384,0.014150222142537435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,3072,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,4096,0.007626666790909237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,3584,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,3584,0.007361777954631382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,2560,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,16384,0.01828977796766493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,3072,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,2048,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,3072,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,2560,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,1536,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,2048,0.006632888896597757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,2560,0.006823110911581252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,1024,0.0046862223082118565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,2048,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,1536,0.00657155571712388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,768,0.0032186667538351486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,1536,0.005975111077229182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,1024,0.00608711110221015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,512,0.0028364445186323593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,1024,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,256,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,6144,0.009361777868535783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,768,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,512,0.005583999885453119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,128,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,512,0.005150222116046482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,256,0.005372444374693765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,64,0.0019004444281260173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,32,0.0021253333737452826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,128,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,256,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,65536,0.03389777739842733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,128,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,2048,0.006863111009200414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,64,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,32,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,16384,0.012130666938092975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,65536,0.04842577709092034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,51200,0.04419555597835117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,12288,0.009659555223253038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,51200,0.038555555873446994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,16384,0.016859556237856548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,10240,0.00868088917599784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,16384,0.017834666702482436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,12288,0.01311288856797748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,8192,0.007672888537247975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,12288,0.014592889282438489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,10240,0.011752000285519494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,10240,0.012423999607563019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,768,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,8192,0.010672888822025724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,6144,0.007068444457319047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,8192,0.009827555881606208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,7168,0.010327111515733931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,5120,0.006234666539563074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,7168,0.009390222529570261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,6144,0.009313777916961247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,4096,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,6144,0.008850666383902231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,3584,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,5120,0.008691555923885768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,5120,0.008027555214034187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,4096,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,65536,0.05478399991989136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,3584,0.007649777664078607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,4096,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,2560,0.004950222041871813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,3584,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,2048,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,2560,0.006986666884687211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,3072,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,1536,0.004559999952713649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,2560,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,2048,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,1024,0.0035706667436493766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,1536,0.006329777754015393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,7168,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,768,0.0031840000301599503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,1536,0.005886222339338726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,1024,0.006266666783226862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,1024,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,768,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,768,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,256,0.0021866667601797315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,512,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,51200,0.027126222848892212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,128,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,256,0.005576888720194499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,3072,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,64,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,256,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,128,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,32,0.001800000046690305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,3072,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,128,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,65536,0.024029332730505202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,64,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,51200,0.019557333654827543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,32,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,65536,0.0468408895863427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,16384,0.01351999988158544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,51200,0.03752266698413425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,51200,0.04378666811519199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,12288,0.011049778097205691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,512,0.002556444456179937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,16384,0.015807110402319167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,16384,0.016512889001104567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,10240,0.010037333601050908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,12288,0.013120888835854001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,8192,0.009365333451165093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,12288,0.0131217779384719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,512,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,10240,0.01205599970287747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,7168,0.008250666989220513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,10240,0.011573333707120685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,6144,0.007082666787836287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,8192,0.010584000084135266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,7168,0.00981422182586458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,8192,0.009852444132169088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,5120,0.006317333214812809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,7168,0.009369778136412302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,6144,0.009334222310119206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,4096,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,5120,0.008616000413894653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,6144,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,3584,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,4096,0.008200888832410177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,3072,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,2048,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,4096,0.007438222567240397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,3584,0.007320000065697565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,2560,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,3584,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,3072,0.007283555964628856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,2048,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,2560,0.006811555474996567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,3072,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,1536,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,2560,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,2048,0.006609777609507243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,1024,0.003871110992299186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,2048,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,1536,0.005971555494599872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,768,0.0032631111227803757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,1024,0.005998222364319696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,1536,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,512,0.0028684443483750024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,1024,0.00564533347884814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,768,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,256,0.002507555608948072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,768,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,512,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,128,0.002151111140847206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,256,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,64,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,256,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,128,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,65536,0.018764444523387484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,128,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,5120,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,65536,0.0543502238061693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,64,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,16384,0.014481777946154276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,32,0.004671999977694618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,65536,0.054125332170062594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,12288,0.010210666391584609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,65536,0.04600266615549723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,51200,0.0429004430770874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,10240,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,16384,0.015799111790127225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,51200,0.036883556180530123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,16384,0.0147724449634552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,12288,0.013109332985348172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,8192,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,7168,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,12288,0.012082666986518435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,8192,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,10240,0.011051555474599203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,6144,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,7168,0.009703111317422655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,5120,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,7168,0.009306666751702627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,6144,0.009337777892748514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,4096,0.005868444426192178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,6144,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,512,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,5120,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,3584,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,5120,0.008004444340864817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,4096,0.00794044468137953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,3584,0.007948444121413762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,4096,0.007535999847782983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,3584,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,51200,0.015789333317014907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,2560,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,3072,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,2048,0.004237333519591225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,3072,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,2560,0.007028444773621029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,1536,0.004559111264016894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,2560,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,2048,0.0069146665434042616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,2048,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,1024,0.003519999898142285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,10240,0.011720000041855706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,768,0.0031724443866146933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,1536,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,1024,0.006299555715587404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,8192,0.009711110757456886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,512,0.0025351111673646504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,1024,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,768,0.005879999862776862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,256,0.002193777718477779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,768,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,128,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,256,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,512,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,64,0.0018497777895794974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,256,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,3072,0.005311111195219888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,768,32,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,128,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,65536,0.02688088847531213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,128,0.004611555486917496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,51200,0.022626666559113398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,64,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,32,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,65536,0.0492142207092709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,16384,0.009335999687512716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,65536,0.045479999648200146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,12288,0.007532444265153673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,51200,0.038726223839653864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,16384,0.01569422251648373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,10240,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,16384,0.014463999205165438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,12288,0.013058666553762225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,1536,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,8192,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,12288,0.012074666718641916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,10240,0.011866666376590729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,8192,0.01055111073785358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,10240,0.010971555279360877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,8192,0.009756444229020013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,7168,0.010047111246320937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,512,0.005650666852792104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,5120,0.00554666668176651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,7168,0.009306666751702627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,4096,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,6144,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,5120,0.008660444782839881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,6144,0.0087013335691558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,3584,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,4096,0.008356444537639618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,3072,0.00535733335547977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,4096,0.007535999847782983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,3584,0.007663999994595845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,2560,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,3584,0.007028444773621029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,3072,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,2048,0.004598222259018156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,3072,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,2560,0.006714666469229593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,1536,0.0035413333939181436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,51200,0.03645688957638211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,2560,0.006600889066855113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,2048,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,2048,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,1536,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,7168,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,768,0.0028275555620590844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,1536,0.005616888817813661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,512,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,1024,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,6144,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,1024,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,768,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,512,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,768,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,128,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,256,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,512,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,256,0.005022222383154763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,128,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,5120,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,65536,0.016015110744370353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,64,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,32,0.004592000196377437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,51200,0.014460444450378418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,65536,0.04419022136264377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,1024,0.0029680000411139596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,51200,0.036981334288915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,51200,0.03427911135885451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,256,0.0022311111291249595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,16384,0.01425866617096795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,12288,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,16384,0.015542222393883599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,64,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,10240,0.00556355549229516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,512,32,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,12288,0.01291822228166792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,8192,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,128,0.004597333156400257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,10240,0.011788444386588203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,7168,0.0053119998839166425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,10240,0.01073688848151101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,8192,0.010520888699425591
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,6144,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,8192,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,65536,0.046226667033301465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,7168,0.009719111025333405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,5120,0.004785777793990241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,7168,0.009358222285906473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,6144,0.009282666775915358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,4096,0.004552000098758274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,5120,0.008519111408127679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,6144,0.008636444807052612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,16384,0.00627377794848548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,3584,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,5120,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,4096,0.008040889269775814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,3584,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,4096,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,2560,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,3072,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,2048,0.004592888885074192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,3584,0.007161777880456712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,3072,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,2560,0.007132444116804335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,1536,0.003536000019974179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,2560,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,2048,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,2048,0.006263111200597551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,12288,0.012088889049159156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,768,0.0026337777574857077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,1536,0.005886222339338726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,1024,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,512,0.002493333278430833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,1024,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,768,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,256,0.0021048889805873237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,768,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,512,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,128,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,512,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,256,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,64,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,256,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,32,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,128,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,65536,0.011381333072980246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,128,0.004605333424276776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,64,0.00453599997692638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,51200,0.009646221995353699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,32,0.004858666823969947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,65536,0.04594488938649496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,16384,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,65536,0.04190488987498813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,12288,0.005233777893914117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,1024,0.002850666642189026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,51200,0.036981334288915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,1536,0.006054222169849608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,16384,0.015811555915408664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,51200,0.033599111768934466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,8192,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,16384,0.014144889182514615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,12288,0.012826667063766055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,7168,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,10240,0.011761777930789523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,8192,0.01075644459989336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,10240,0.010746666954623329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,6144,0.004593777573770947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,8192,0.009673777553770278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,5120,0.004561777744028303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,7168,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,6144,0.009085333181752099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,4096,0.004261333495378494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,6144,0.008481777376598781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,5120,0.00849511143234041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,5120,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,4096,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,4096,0.0069884442620807225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,3072,0.004250666747490565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,3584,0.007371555599901412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,3584,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,2560,0.00452977791428566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,3072,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,3072,0.006800000038411882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,10240,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,2048,0.004234666625658671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,2560,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,2560,0.006284444282452266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,12288,0.012311999996503195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,1536,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,2048,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,1024,0.0028648889727062653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,1536,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,1024,0.006301333506902059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,1024,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,768,0.002565333412753211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,3584,0.004651555584536659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,512,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,768,0.005984000033802456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,768,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,256,0.002128888956374592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,512,0.00562755556570159
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,512,0.0052728888889153796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,128,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,256,0.005632888939645555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,256,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,128,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,128,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,2048,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,32,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,65536,0.006721777634488211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,1536,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,51200,0.0069120000633928515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,64,0.004931555440028508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,32,0.004639111045334074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,16384,0.005364444520738389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,12288,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,65536,0.04173422190878126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,10240,0.004577777865860197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,256,3072,0.004643555730581284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,8192,0.0047173334492577445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,51200,0.03348711133003235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,7168,0.004565333326657613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,16384,0.014161777165200977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,6144,0.004554666578769684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,10240,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,5120,0.004599110947714912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,8192,0.009373333719041612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,4096,0.003951999876234266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,7168,0.008957333034939235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,3584,0.0041706665522522396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,6144,0.008303110798199972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,3072,0.0042364444169733255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,5120,0.007987555530336168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,2560,0.004560889055331548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,4096,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,128,64,0.0018079999006456798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,2048,0.0042239998777707415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,3584,0.0069057775868309875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,1536,0.0035573335157500375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,1024,0.0028640000770489373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,3072,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,768,0.002483555633160803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,7168,0.010007999837398529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,2560,0.006309333360857434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,512,0.0024888888001441956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,256,0.001863111017478837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,1536,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,128,0.0018373332503769135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,1024,0.005260444349712796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,64,0.0015333333156175083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,768,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,64,32,0.00150222217457162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,65536,0.006649777707126405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,256,0.005054222212897407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,51200,0.006594666590293248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,128,0.00453244439429707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,12288,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,16384,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,32,0.004599110947714912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,12288,0.005186666631036335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,10240,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,65536,0.041385776466793485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,8192,0.0041884444653987885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,51200,0.03377066718207465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,16384,0.014477334088749357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,7168,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,6144,0.004545777622196409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,10240,0.01071911139620675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,5120,0.0038480001191298165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,8192,0.0096142225795322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,4096,0.003843555433882607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,7168,0.009000000026490953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,6144,0.008639111287064022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,5120,0.007603555917739868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,2048,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,3072,0.003855111284388436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,4096,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,2560,0.004223111189073986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,2048,0.003875555677546395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,3584,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,1536,0.0035004446076022256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,512,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,1024,0.00286311118139161
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,3072,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,768,0.0024702221983008915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,2560,0.006302222195598815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,2048,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,512,0.0021679999513758556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,1536,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,256,0.0019217777169413036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,1024,0.00527555536892679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,128,0.0018311111877361934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,768,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,64,0.0017964444640609955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,512,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,12288,0.01182666669289271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,256,0.00507733349998792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,128,0.004932444542646408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,64,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,3584,0.004559111264016894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,64,0.004624888714816836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,51200,0.4854577912224664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,51200,1.093967119852702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,16384,0.33393155203925234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,5,32,32,0.001485333364042971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,16384,0.18137689431508383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,12288,0.2559751139746772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,12288,0.15123466650644937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,51200,0.43245866563585067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,16384,0.13438933425479466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,10240,0.21300621827443442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,12288,0.10230222013261582
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,10240,0.13568088743421766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,8192,0.16539822684393987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,10240,0.08751377794477676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,8192,0.11890489525265163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,32,0.004596444467703502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,8192,0.07110933462778728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,7168,0.14351644780900744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,10240,0.11162666479746501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,6144,0.1240239938100179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,16384,0.1737582286198934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,7168,0.06093066930770874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,6144,0.05336888962321811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,7168,0.08253955841064453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,5120,0.10198489162656997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,6144,0.0721528861257765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,12288,0.1302373276816474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,5120,0.07997955878575642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,4096,0.08481066756778294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,7168,0.10634666681289673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,5120,0.04645688997374641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,8192,0.0942186646991306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,3584,0.07220711310704549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,5120,0.06273866362041898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,4096,0.03896355628967285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,3072,0.06203999784257677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,4096,0.06641155481338501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,3584,0.05464533302519056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,4096,0.05272710985607571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,6144,0.08891733487447102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,2560,0.052453332477145724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,3072,0.05323644479115804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,3072,0.031256887647840716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,3584,0.04790488878885905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,2048,0.04243555665016174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,2560,0.040207111173205905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,2560,0.028806222809685603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,1536,0.03203466534614563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,3072,0.04333422250217862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,2048,0.03622400098376804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,2048,0.028020444843504164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,2560,0.03759288787841797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,1024,0.02267644471592373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,1536,0.02899822261598375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,51200,0.5035022099812826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,768,0.017863111363516915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,2048,0.03360533383157518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,1024,0.018060444129837882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,1536,0.028558221128251817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,1024,0.021135111649831135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,512,0.012679999901188744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,768,0.01515111161602868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,3584,0.034182220697402954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,1024,0.02288355595535702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,256,0.007640889121426477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,512,0.012780444489585029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,768,0.01976355579164293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,512,0.013072888884279462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,128,0.005203555441564984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,256,0.01036266651418474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,512,0.017755554782019723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,256,0.010293333066834344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,64,0.004224888980388641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,32,0.0042088888585567474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,128,0.010499555203649731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,256,0.014396443963050842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,128,0.007962666451931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,64,0.00903200027015474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,128,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,32,0.009017777939637503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,1536,0.023246222072177466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,768,0.016918222109476726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,65536,0.515176879035102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,51200,0.3885964552561442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,51200,0.7888897789849175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,65536,1.0065253575642903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,16384,0.25621244642469615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,16384,0.1467342244254218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,16384,0.10567733314302231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,12288,0.19041777981652153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,65536,0.44274044036865234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,12288,0.12436177995469834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,51200,0.36159467697143555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,10240,0.1584951082865397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,12288,0.08301333586374919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,10240,0.10336977905697292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,8192,0.12947200404273138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,10240,0.07724799712498982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,12288,0.10562844408882989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,16384,0.14254666699303523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,8192,0.09249155388938056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,7168,0.11338399516211616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,10240,0.09055733018451267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,8192,0.05765955315695869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,7168,0.08192711406283908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,6144,0.0986586676703559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,8192,0.07629777987798055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,7168,0.04976888828807407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,5120,0.08833066622416179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,6144,0.07269155316882663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,6144,0.046965334150526256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,7168,0.06816177897983126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,4096,0.07123644484413995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,5120,0.03876533442073398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,6144,0.06035377581914266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,5120,0.06543199883566962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,3584,0.06255377663506402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,4096,0.054766221178902515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,4096,0.03257511059443156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,5120,0.05202577842606438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,4096,0.04412710997793409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,3072,0.054506666130489774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,3584,0.044101334280437894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,3584,0.02915289004643758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,2560,0.04566755559709337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,3072,0.04381777842839559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,3072,0.025968889395395916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,3584,0.04199644592073229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,2048,0.03684355484114753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,2560,0.03304266598489549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,2560,0.024135111106766596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,1536,0.028453333510292902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,3072,0.03653511073854234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,2048,0.023306666149033442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,2560,0.03420622150103251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,2048,0.030808889203601416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,1024,0.020290666156344943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,1536,0.023260444402694702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,1536,0.01978133287694719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,768,0.014860444598727755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,2048,0.029790222644805908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,1536,0.024661334024535284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,1024,0.017875555488798354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,512,0.010535111029942831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,768,0.013406222065289816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,1024,0.020112888680564035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,512,0.01109866632355584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,768,0.017875555488798354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,512,0.011391999820868174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,256,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,51200,0.3928702142503526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,128,0.0031582222630580268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,256,0.009377777576446533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,512,0.0159608887301551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,256,0.008692444198661381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,64,0.0025111111915773815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,128,0.009301332963837517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,128,0.007346666521496243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,256,0.012488000094890594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,32,0.0028159999185138275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,64,0.008266666697131263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,128,0.010086221827401055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,32,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,1024,0.016006223029560514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,65536,0.319651550716824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,65536,0.17458844184875488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,65536,0.5160924593607584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,768,0.014522666732470194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,51200,0.2481537659962972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,16384,0.08438044124179417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,65536,0.14431733555263943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,51200,0.13030221727159289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,12288,0.06486666864818998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,16384,0.06679999828338623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,51200,0.13334222634633383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,16384,0.041915555795033775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,10240,0.055028445190853544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,12288,0.05115822288725111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,12288,0.034006221426857844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,8192,0.041880001624425255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,10240,0.046521776252322726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,16384,0.05653422408633762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,10240,0.02952622373898824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,12288,0.044440888696246676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,7168,0.03604977660708957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,8192,0.024496000674035814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,10240,0.03767377800411648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,8192,0.03678311241997613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,6144,0.03141155507829454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,7168,0.03327733278274536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,7168,0.02650044361750285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,8192,0.03417777683999803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,51200,0.13640799787309435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,5120,0.027718222803539697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,6144,0.02027555637889438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,6144,0.02989333205752903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,7168,0.03190133306715224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,65536,0.1744542254341973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,4096,0.023074666659037273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,5120,0.026143110460705225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,6144,0.02879200047916836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,5120,0.017495999733606975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,3584,0.019543111324310303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,4096,0.016227554943826463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,4096,0.022388445006476507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,3072,0.017238222890430026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,3584,0.014192889134089151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,3584,0.018955555227067735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,5120,0.025342222717073228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,2560,0.014712888333532544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,4096,0.022330666581789654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,3072,0.013080000049538083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,2048,0.012562666502263812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,3584,0.020280000236299302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,2560,0.01276533305644989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,3072,0.01942488882276747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,1536,0.009676444033781687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,2048,0.012969777815871768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,2560,0.018246221873495314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,2048,0.013832888669437833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,2048,0.016547555724779766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,1024,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,1536,0.01221244451072481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,1536,0.01076533314254549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,768,0.00582755563987626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,1024,0.009004444711738164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,1536,0.015136889285511441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,1024,0.009658666948477427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,512,0.00416711096962293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,768,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,768,0.008062221937709385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,512,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,512,0.006719999843173557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,256,0.002537777854336633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,3072,0.01807733376820882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,512,0.011666666302416058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,256,0.006599111275540457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,2560,0.015278221832381355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,256,0.0059164443777667145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,128,0.006754666566848755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,64,0.0021253333737452826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,128,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,32,0.002019555618365606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,128,0.00870577742656072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,64,0.005912000106440649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,32,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,1024,0.013464888764752282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,65536,0.24263999197218153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,65536,0.14806044101715088
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,51200,0.1878960000144111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,65536,0.1292862229877048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,51200,0.11139733261532253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,128,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,16384,0.0658408866988288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,256,0.010728889041476779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,51200,0.10817955599890815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,16384,0.03727555606100295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,12288,0.04953155583805508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,16384,0.05891022417280409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,10240,0.04215111003981697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,16384,0.04466133316357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,12288,0.030010667112138536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,12288,0.045309331681993276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,8192,0.036944889360004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,10240,0.040425777435302734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,10240,0.025841777523358662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,12288,0.03567911187807719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,65536,0.1363324456744724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,7168,0.03277511066860623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,51200,0.10565777619679768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,8192,0.03228266702757941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,768,0.011739555332395764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,10240,0.032187554571363665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,6144,0.028720001379648846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,7168,0.022869333624839783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,7168,0.028759111960728962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,5120,0.024664888779322307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,6144,0.017872888180944655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,6144,0.026477333572175767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,7168,0.02630311085118188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,4096,0.02112355497148302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,5120,0.015871110889646742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,5120,0.02257688840230306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,6144,0.02401155564520094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,3584,0.018798222144444782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,4096,0.014196443888876172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,4096,0.01997155613369412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,5120,0.02104355560408698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,3072,0.016397333807415433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,3584,0.016688000824716356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,4096,0.019511111908488803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,3584,0.013393777940008374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,2560,0.013929777675204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,3072,0.012413333687517377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,3584,0.018179555733998615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,8192,0.021872889664438035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,2560,0.01384800010257297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,2048,0.012431999875439538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,2560,0.01143200033240848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,8192,0.029091556866963703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,2560,0.016557332542207506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,1536,0.007947555846638149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,2048,0.013061333033773633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,1536,0.009928888744778102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,1024,0.006467555546098285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,1536,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,1536,0.014096000128322177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,1024,0.008630221916569604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,768,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,1024,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,1024,0.012460444536474017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,3072,0.016698666744761996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,512,0.003112000102798144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,768,0.0074026667409473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,3072,0.01809599995613098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,768,0.010678222609890832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,256,0.002473777780930201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,512,0.0069955554273393415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,256,0.006487999939256244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,512,0.010323555933104621
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,256,0.005608888963858287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,2048,0.015498666299713982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,256,0.010096888575288985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,64,0.0019173332386546666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,128,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,32,0.002128888956374592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,128,0.008353778057628209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,64,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,32,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,65536,0.20248266061147055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,65536,0.1391413344277276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,2048,0.011256888508796692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,51200,0.15816444820827907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,65536,0.12502488825056288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,128,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,16384,0.056484447585211865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,51200,0.10653422276178996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,128,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,51200,0.10023733642366196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,16384,0.048130667871899076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,12288,0.0428071121374766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,16384,0.03486488925086127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,65536,0.11968444453345405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,16384,0.0403128895494673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,12288,0.02755911151568095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,12288,0.037224001354641385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,12288,0.03223911258909438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,10240,0.024312888582547505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,512,0.006573333508438534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,8192,0.03043466806411743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,10240,0.029666665527555678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,8192,0.02081155611409081
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,7168,0.026917333404223125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,8192,0.027069333526823256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,7168,0.018826666805479262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,8192,0.0259253333012263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,7168,0.023579556081030104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,7168,0.024018666810459558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,6144,0.021388444635603163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,5120,0.02095555596881443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,6144,0.01726222203837501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,10240,0.03703644540574815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,6144,0.02188266648186578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,4096,0.017834666702482436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,51200,0.09330311086442734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,10240,0.031910220781962075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,4096,0.013782222237851886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,4096,0.016586666305859882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,5120,0.019984000258975558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,3584,0.015791111522250705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,3584,0.012800888882742988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,4096,0.01863555610179901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,6144,0.024684444069862366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,3584,0.014340443743599786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,3584,0.017902221944597032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,3072,0.011744000017642975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,2560,0.012446222205956778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,3072,0.014045332868893942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,768,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,3072,0.017180444465743173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,2560,0.01072355525361167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,2048,0.009557333257463243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,2560,0.012327999704413943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,1536,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,5120,0.01940088967482249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,2048,0.010038221875826517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,2560,0.01575377749072181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,2048,0.011678222152921887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,1024,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,1536,0.008994667066468133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,1536,0.009015111459626092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,2048,0.015378667248619927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,768,0.0038311113086011675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,1024,0.008001777860853408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,1024,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,1536,0.014478221535682678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,512,0.0032106666929192017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,3072,0.01405244403415256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,768,0.0069022224181228215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,768,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,1024,0.011358222199810876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,512,0.006909333169460297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,512,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,768,0.011053333679835001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,128,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,256,0.00554933316177792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,512,0.010052444206343757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,5120,0.017136888371573556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,128,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,128,0.006308444258239534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,256,0.010388444695207808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,32,0.0019235555082559586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,128,0.007956444389290279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,65536,0.16820532745785186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,64,0.005659555395444234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,32,0.005571555760171678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,51200,0.12710132863786486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,65536,0.08790400293138291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,65536,0.07130222188101874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,256,0.002437333265940348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,16384,0.045279999574025474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,51200,0.06863200002246432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,256,0.00658044425977601
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,51200,0.05854666895336575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,12288,0.035269333256615534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,16384,0.035762667655944824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,16384,0.022221333450741235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,64,0.0019448887970712448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,65536,0.11004088984595405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,10240,0.02882222334543864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,12288,0.01834755473666721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,12288,0.027793778313530818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,51200,0.08862488799624973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,8192,0.022201778160201177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,12288,0.03192889028125339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,10240,0.016139555308553908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,10240,0.02438666588730282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,7168,0.019840000404251944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,8192,0.021382222572962444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,8192,0.014451555079884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,10240,0.028459555572933618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,6144,0.017465778523021273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,7168,0.013425778183672162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,5120,0.015018666783968607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,7168,0.018592000007629395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,8192,0.024829333027203877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,6144,0.012402666939629449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,4096,0.013472889032628802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,6144,0.01741155485312144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,5120,0.011731555892361535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,5120,0.015848888291252982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,3584,0.011199110911952125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,6144,0.02146666745344798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,4096,0.010558221903112201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,4096,0.01334755536582735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,5120,0.019395555059115093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,3072,0.01035644445154402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,3584,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,4096,0.018611555298169453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,3584,0.012267555627557965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,2560,0.008604444563388824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,16384,0.03797422183884515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,3072,0.009003555609120263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,2048,0.007711110843552484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,3584,0.01718755563100179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,2560,0.008272888759771982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,2560,0.010494222243626913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,3072,0.016135111451148987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,1536,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,2048,0.00793422261873881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,2048,0.009568000005351173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,2560,0.01625955601533254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,1024,0.004553777890072929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,1536,0.007252444823582967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,1536,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,768,0.003168888803985384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,7168,0.022329777479171753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,1024,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,1024,0.006656000183688269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,1536,0.013103110922707452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,512,0.0025235555238193935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,768,0.006265777680608962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,1024,0.01148977792925305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,768,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,256,0.0021964444054497615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,512,0.006093333164850871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,512,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,768,0.010743111371994019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,128,0.0021342221233579847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,256,0.005220444252093633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,256,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,3072,0.01145511120557785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,64,0.0019173332386546666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,128,0.005919999960396025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,32,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,256,0.010379555324713389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,64,0.0053191110491752625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,65536,0.14669777287377253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,32,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,2048,0.014807111687130399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,51200,0.11464888519710964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,65536,0.08227555619345771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,65536,0.06594844659169515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,51200,0.06207378043068779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,51200,0.05346044566896227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,16384,0.040748443868425153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,512,0.011071110765139261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,65536,0.10998133156034683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,16384,0.02160177793767717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,128,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,12288,0.032664888434939914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,16384,0.03368533319897122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,51200,0.0886497762468126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,10240,0.02724444369475047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,128,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,12288,0.01769333415561252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,16384,0.03762755460209317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,12288,0.027092445227834914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,8192,0.023810666468408372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,10240,0.015944000747468736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,10240,0.023400889502631292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,7168,0.01776622235774994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,8192,0.014017777310477363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,12288,0.030897776285807293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,6144,0.015867556134859722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,10240,0.027971555789311726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,8192,0.02033866610791948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,7168,0.01293155550956726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,7168,0.01828977796766493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,8192,0.024528000089857314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,6144,0.01202933324707879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,6144,0.016773333152135212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,7168,0.022333333889643352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,4096,0.0123831108212471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,5120,0.010935111178292168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,5120,0.014807111687130399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,6144,0.021533333592944678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,3584,0.010294222169452244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,4096,0.009684444301658207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,5120,0.019839111301634047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,4096,0.012712889247470431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,3072,0.00943377779589759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,3584,0.008973333570692275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,3584,0.011707555916574268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,4096,0.017601778109868366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,2560,0.008107555409272512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,3072,0.007984889050324758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,3072,0.011015111373530494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,3584,0.016700444949997794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,2048,0.007010666860474481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,2560,0.00980355590581894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,3072,0.016580444243219163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,2048,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,2048,0.0092275556590822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,2560,0.01581333412064446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,5120,0.013640888863139682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,1024,0.003942222230964237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,1536,0.007592000067234039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,2048,0.014420444766680399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,768,0.0030622221529483795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,1024,0.006312888943486744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,1024,0.006585777633719974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,1536,0.01239911135700014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,512,0.002534222271707323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,1024,0.011756444142924415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,512,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,768,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,256,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,512,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,512,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,2560,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,128,0.002096000024014049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,256,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,256,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,128,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,256,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,128,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,64,0.0018586667461527719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,32,0.0018240000224775737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,65536,0.1212728950712416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,128,0.008719111482302347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,64,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,32,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,768,0.006092444476154115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,51200,0.09717333316802979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,65536,0.07462666432062785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,51200,0.056073778205447726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,1536,0.0057653333577844836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,16384,0.033687111404207014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,65536,0.10978755686018203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,1536,0.00664444433318244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,51200,0.08852710988786484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,16384,0.031269331773122154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,12288,0.02793955471780565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,16384,0.02091911103990343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,16384,0.03731733229425218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,768,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,10240,0.024159111910396155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,12288,0.02511288887924618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,12288,0.017218665944205392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,8192,0.020255999432669747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,10240,0.01516177753607432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,10240,0.021990221407678392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,12288,0.03086044390996297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,7168,0.017803554733594257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,8192,0.01923288901646932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,10240,0.02736888825893402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,8192,0.013733333183659447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,6144,0.01610933409796821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,51200,0.05134488807784187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,7168,0.012514666550689273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,8192,0.0237128883600235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,7168,0.017478222648302715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,5120,0.013813333378897773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,6144,0.011614222493436603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,6144,0.015779554843902588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,7168,0.022132444712850783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,4096,0.010955555571450127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,5120,0.010407110883129967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,5120,0.01421333352724711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,6144,0.020695999264717102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,3584,0.009298666483826106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,4096,0.012771555946932899
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,4096,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,3072,0.008404444489214156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,3584,0.00868800034125646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,3584,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,4096,0.017866666118303936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,65536,0.06321333514319526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,2560,0.007263999846246507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,3072,0.01075466639465756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,3072,0.007657777931955125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,3584,0.016978666186332703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,2048,0.006350222147173352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,2560,0.0074675555030504865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,2560,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,1536,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,3072,0.015722667177518208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,2560,0.015433778365453085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,2048,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,1024,0.003495111233658261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,1536,0.006703111032644908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,2048,0.013077333569526672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,1536,0.0070186663005087115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,768,0.0028284444577164124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,1024,0.00630666646692488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,1536,0.011909333368142446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,512,0.002540444541308615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,768,0.005954666684071223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,1024,0.01143022212717268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,5120,0.018932445181740653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,256,0.002187555655837059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,512,0.0058782220714622075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,512,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,256,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,256,0.005659555395444234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,128,0.0021466666625605687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,512,0.010211555494202508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,128,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,256,0.01036888857682546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,64,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,128,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,2048,0.008405333591832055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,128,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,65536,0.1054275565677219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,64,0.005224888937340842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,32,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,1024,0.006565333240562015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,51200,0.08215644624498156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,65536,0.06668621963924833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,768,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,65536,0.0608231094148424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,16384,0.03132088979085287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,51200,0.05058844553099739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,768,0.010372444159454769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,12288,0.02542488939232296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,51200,0.049294223388036094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,16384,0.02995022137959798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,16384,0.020206222931543987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,10240,0.021176000436147053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,65536,0.10936888721254136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,51200,0.08826133277681138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,16384,0.03691555394066705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,8192,0.01854222185081906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,32,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,10240,0.014833778142929077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,10240,0.020305777589480083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,8192,0.017857778403494094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,8192,0.013079110946920184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,10240,0.027021333575248718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,7168,0.01626133256488376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,7168,0.012178666889667511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,8192,0.023336889015303716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,6144,0.014520888527234396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,7168,0.016176000237464905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,6144,0.01108000013563368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,7168,0.02218666672706604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,5120,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,6144,0.01514755520555708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,5120,0.009751111268997192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,4096,0.010025777750545079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,12288,0.016765332884258695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,5120,0.01924888955222236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,12288,0.029706666866938274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,3584,0.00849777791235182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,4096,0.012440889245933957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,4096,0.008665777742862701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,4096,0.017828444639841717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,3072,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,3584,0.008062221937709385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,3584,0.010438222024175854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,3072,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,3584,0.01646399994691213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,2560,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,3072,0.010027555955780877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,2048,0.006046222315894232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,2560,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,2560,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,6144,0.020976000361972388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,1536,0.0045422220395671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,2048,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,5120,0.013429332938459186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,2048,0.007717333734035492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,1024,0.003528000166018804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,1536,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,1536,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,2048,0.013169777890046438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,768,0.0028417776856157514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,1024,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,1024,0.006245333287451003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,1536,0.011682666838169098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,512,0.0025235555238193935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,768,0.005967111223273807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,768,0.005715555614895291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,1024,0.011074666347768573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,256,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,512,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,512,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,768,0.011047110789351992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,128,0.0019137778629859288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,3072,0.016146666473812528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,512,0.010850666297806634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,128,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,12288,0.023936000135209825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,64,0.001869333287080129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,256,0.01035822182893753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,32,0.0018871112002266778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,65536,0.08750222126642863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,128,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,64,0.00510044437315729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,32,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,51200,0.06850577725304498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,65536,0.06078400214513143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,65536,0.05866310993830363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,51200,0.0461715559164683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,65536,0.10936355590820312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,16384,0.02683199942111969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,256,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,256,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,51200,0.08816711107889812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,12288,0.01902044481701321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,16384,0.02586488922437032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,16384,0.01942755613062117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,2560,0.013777777552604675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,10240,0.016893333858913846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,12288,0.016159110599093966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,16384,0.03637511200375027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,12288,0.021482666333516438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,8192,0.01513155632548862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,10240,0.014282666974597506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,10240,0.020245333512624104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,7168,0.013399110900031196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,12288,0.029622223642137315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,8192,0.012501333322789935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,10240,0.02607466611597273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,8192,0.016927111479971144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,6144,0.011743110915025076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,7168,0.016140444411171805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,7168,0.011396444506115384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,8192,0.02367288867632548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,5120,0.01069866700304879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,51200,0.04748444424735176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,6144,0.010390222072601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,7168,0.021900445222854614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,5120,0.009134222235944536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,5120,0.012432888978057437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,5120,0.01908533275127411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,128,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,4096,0.011350221931934357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,3584,0.01032977799574534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,4096,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,4096,0.01732711162832048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,3072,0.009005332986513773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,3584,0.008005333443482717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,3584,0.009860444400045607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,3584,0.01680355601840549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,3072,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,3072,0.00924444446961085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,3072,0.014847111370828418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,6144,0.013958222336239286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,2048,0.007003555695215861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,2560,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,2560,0.007296000089910295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,6144,0.020303999384244282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,1536,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,2560,0.013734222286277346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,2048,0.006768888897365994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,2048,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,1024,0.0042239998777707415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,1536,0.006663110935025745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,1536,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,2048,0.01351288871632682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,768,0.0038319999972979226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,1024,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,1536,0.012076444096035428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,1024,0.005984000033802456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,512,0.0032008888406885993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,768,0.005941333456171884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,768,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,2560,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,256,0.002193777718477779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,512,0.005377777748637729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,512,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,768,0.010406222608354356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,128,0.0019173332386546666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,256,0.005618666609128316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,4096,0.009507555928495195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,512,0.010666666759385003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,64,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,128,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,256,0.010029333333174387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,32,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,65536,0.07974577612347074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,64,0.005200888961553574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,128,0.008077333370844523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,32,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,51200,0.06132266918818156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,65536,0.05674755573272705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,65536,0.057842665248446994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,16384,0.023846222294701472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,1024,0.011436444189813403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,51200,0.04668177829848396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,51200,0.04404533240530226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,65536,0.1090186635653178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,12288,0.018266666266653273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,16384,0.019288000133302476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,16384,0.025228444072935317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,51200,0.08778755532370673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,256,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,10240,0.015140444040298462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,12288,0.01924444403913286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,12288,0.015825778245925903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,16384,0.036193778117497764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,8192,0.013399110900031196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,128,0.004962666581074397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,10240,0.014055111342006259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,7168,0.012429333395428128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,12288,0.029470221863852605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,10240,0.018587556150224473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,8192,0.01220533334546619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,6144,0.01110755569405026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,8192,0.016492444607946608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,7168,0.011423110961914062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,10240,0.026725333597924974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,5120,0.009687110781669617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,7168,0.014820444915029736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,6144,0.009699555734793345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,8192,0.02366399930583106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,4096,0.008685333033402761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,6144,0.012650666965378655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,7168,0.021648888786633808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,5120,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,3584,0.009639999932712978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,5120,0.01239999963177575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,6144,0.020076443751653034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,4096,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,4096,0.01073066641887029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,3072,0.00850666645500395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,5120,0.018782221608691745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,3584,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,2560,0.00795377790927887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,3584,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,4096,0.016916443904240925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,3072,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,2048,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,3072,0.007891555627187094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,3584,0.015150222513410779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,2560,0.007415999968846639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,2560,0.007618666523032718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,3072,0.015423110789722867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,2048,0.006715555571847492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,2048,0.0070462218589252895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,2560,0.01349422252840466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,1536,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,1536,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,2048,0.012784000072214337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,768,0.0035128887328836652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,1024,0.0063031112982167145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,1536,0.012095111111799875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,768,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,768,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,512,0.0028408887899584244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,1024,0.01072088877360026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,768,0.01109333336353302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,256,0.002149333349532551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,512,0.010377777947319878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,1536,0.005117333183685939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,256,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,128,0.0018613334331247543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,256,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,128,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,256,0.01036800030204985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,1024,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,64,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,32,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,128,0.005640000104904175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,128,0.008038221961922115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,65536,0.06801422437032063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,64,0.004962666581074397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,32,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,51200,0.05198577708668179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,512,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,512,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,65536,0.05392000079154968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,65536,0.05683288971583048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,16384,0.019807111885812547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,51200,0.045808888143963285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,51200,0.04239733351601494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,65536,0.10904533333248562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,12288,0.016388444436921012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,16384,0.023917333947287664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,1024,0.0038951109680864546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,16384,0.01889955500761668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,10240,0.013910222384664746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,51200,0.08780088689592148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,12288,0.015641777051819693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,8192,0.012093333734406365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,12288,0.018365333477656048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,16384,0.03566311134232415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,10240,0.013794666363133324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,10240,0.016194666425387066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,7168,0.011221333510345884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,12288,0.029614223374260798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,8192,0.01204622205760744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,6144,0.010295111272070143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,8192,0.015331554743978711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,10240,0.026166222161716882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,7168,0.01090311093462838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,5120,0.008967999782827165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,7168,0.013207111093733044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,8192,0.023374223046832617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,6144,0.0099973330895106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,6144,0.012467555701732635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,7168,0.021344888541433547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,5120,0.008799110849698385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,5120,0.01107022249036365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,6144,0.020246222615242004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,3584,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,4096,0.008248888783984715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,5120,0.018595554762416415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,3072,0.008620444271299574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,3584,0.007931555310885111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,3584,0.008037333687146505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,4096,0.016703110602166917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,2560,0.007242666350470648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,3072,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,3072,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,3584,0.015276443627145557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,2048,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,2560,0.0069822221994400024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,2560,0.007119999991522894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,3072,0.014775110615624322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,1536,0.005374222166008419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,2048,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,2560,0.013448889056841532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,1024,0.004212444441186057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,4096,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,1536,0.005964444329341252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,2048,0.012479999827014076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,768,0.0038328886859946777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,1024,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,1536,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,1024,0.005740444279379315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,4096,0.00872622181971868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,512,0.0035075553589397003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,768,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,1024,0.010683555569913654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,512,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,256,0.002833777831660377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,768,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,512,0.009708444277445475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,256,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,2048,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,128,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,256,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,256,0.010692444112565784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,64,0.001810666587617662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,128,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,32,0.0018959999498393799
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,1536,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,128,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,65536,0.058077335357666016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,128,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,64,0.004925333377387789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,32,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,51200,0.04598399996757507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,768,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,65536,0.051608890295028687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,16384,0.018409777018758986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,65536,0.055792000558641225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,512,0.0052933332820733385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,51200,0.044920000765058726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,51200,0.041088001595603094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,12288,0.01424800025092231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,65536,0.1090631087621053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,16384,0.02120977805720435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,16384,0.01885422236389584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,10240,0.012491555677519904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,51200,0.08792355325486924
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,12288,0.01516355574131012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,12288,0.018584888842370775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,8192,0.010821333361996545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,16384,0.03525688913133409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,10240,0.01314311143424776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,10240,0.015632889337009855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,12288,0.029655999607510034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,8192,0.01111822244193819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,6144,0.008935111264387766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,8192,0.013919111755159167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,10240,0.026407111022207472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,7168,0.010332444475756751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,5120,0.008367111285527548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,8192,0.022987556126382615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,7168,0.012784889174832238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,6144,0.009409777820110321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,6144,0.01146755533085929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,4096,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,5120,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,5120,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,6144,0.01958488921324412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,3584,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,4096,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,4096,0.008082666330867344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,5120,0.017807111144065857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,3072,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,3584,0.007667555577225155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,4096,0.016176889340082806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,3584,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,2560,0.006682666639486949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,3072,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,3072,0.007060444189442529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,3584,0.015198222464985318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,2048,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,2560,0.0069422221018208405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,2560,0.006468444648716185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,7168,0.009887110855844285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,1536,0.005931555396980709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,2048,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,2048,0.006227555374304454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,2560,0.013769778112570444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,1024,0.00416088890698221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,1536,0.006579555571079254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,2048,0.012791111237472959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,768,0.0035333335399627686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,1024,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,7168,0.021563554803530376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,1024,0.011431111229790581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,768,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,512,0.002875555513633622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,768,0.01071733319097095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,512,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,256,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,512,0.00565422202150027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,512,0.010656000011497073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,3072,0.0141004439857271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,128,0.0021546667234765156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,256,0.005547555370463266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,256,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,1536,0.005966222120655908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,128,0.00553688903649648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,128,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,32,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,1024,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,1536,0.011744000017642975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,65536,0.048657778236601085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,128,0.008468444148699442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,64,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,768,0.005533333453867171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,32,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,51200,0.039515554904937744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,65536,0.05015822251637777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,65536,0.055126223299238414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,16384,0.014161777165200977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,51200,0.04473155405786303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,51200,0.03973689013057285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,65536,0.10894578033023411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,12288,0.01167377746767468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,51200,0.08760888708962335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,16384,0.01854222185081906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,16384,0.02017155620786879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,10240,0.010680889089902243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,12288,0.014363555444611443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,16384,0.03491466575198703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,64,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,8192,0.009415999882751042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,10240,0.013193777865833707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,10240,0.015091554986106025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,12288,0.029271999994913738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,7168,0.008667555948098501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,8192,0.012056888805495368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,8192,0.011004444625642566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,10240,0.02575911084810893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,6144,0.0076142218377855085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,7168,0.012105777859687805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,7168,0.01035199976629681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,8192,0.022673777408070032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,5120,0.007019555403126611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,6144,0.00964355551534229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,6144,0.01035199976629681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,7168,0.02166044380929735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,4096,0.008616000413894653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,5120,0.008960000342792934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,5120,0.009031111167536842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,6144,0.019347555107540555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,3584,0.008311111066076491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,4096,0.007994666695594788
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,4096,0.007609777980380588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,256,0.009967111051082611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,5120,0.0177848885456721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,3072,0.006341333190600078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,3584,0.007663999994595845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,4096,0.015554666519165039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,2560,0.005197333378924264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,3072,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,3072,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,3584,0.01479822231663598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,2048,0.007625777688291337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,2560,0.006968888971540663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,3072,0.01407999959256914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,2560,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,1536,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,2048,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,2048,0.00618488879667388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,2560,0.013460444079505073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,1024,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,1536,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,1536,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,768,0.0034799998005231223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,2048,0.01272266689274046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,1024,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,1024,0.005624889085690181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,512,0.002826666666401757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,1536,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,12288,0.015976000163290236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,768,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,1024,0.010738666686746808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,768,0.0052720002002186244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,3584,0.007287110719415877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,512,0.005586666779385672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,512,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,768,0.010616000327799054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,256,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,256,0.004964444372389051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,512,0.010777778095669217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,128,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,128,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,256,0.009343999955389235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,65536,0.03585955500602722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,64,0.005071111023426056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,128,0.0083155557513237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,32,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,51200,0.02884622083769904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,65536,0.047817776600519814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,65536,0.054711110062069364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,51200,0.044397334257761635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,51200,0.038655112187067665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,65536,0.10908889108233982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,16384,0.012328888807031842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,256,0.0024968888610601425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,12288,0.00979911122057173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,128,0.0021235555824306277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,16384,0.016677333248986136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,51200,0.08694666624069214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,16384,0.017743999759356182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,10240,0.008728889127572378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,64,0.00184888889392217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,12288,0.014652444256676568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,12288,0.013456000222100152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,8192,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,16384,0.03533688849873013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,32,0.0020373333245515823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,10240,0.011832888755533429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,7168,0.008265777594513362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,12288,0.02914488977856106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,8192,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,8192,0.010001777774757808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,6144,0.007260444263617198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,10240,0.02574133376280467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,7168,0.00998933364947637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,7168,0.009642666412724389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,8192,0.02327555583582984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,5120,0.006640000061856375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,6144,0.009348444640636444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,6144,0.00868888861603207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,7168,0.020909332566791113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,4096,0.006321777900060018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,5120,0.00868977771864997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,5120,0.00831288927131229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,6144,0.018990221950742934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,3584,0.005728888842794631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,4096,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,4096,0.007450666692521837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,5120,0.01718133356836107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,3072,0.005532444351249271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,3584,0.007672000262472365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,3584,0.007300444775157505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,2560,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,4096,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,3072,0.007333333293596904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,3072,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,3584,0.015207999282413058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,10240,0.012450666891203986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,2560,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,3072,0.014480888843536377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,2560,0.0064195555945237475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,1536,0.004688888788223267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,2048,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,2560,0.013427555561065674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,1024,0.0036151111125946045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,1536,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,1536,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,1536,0.01203733351495531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,1024,0.006340444501903322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,768,0.003152888889114062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,1024,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,768,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,1024,0.011411555939250521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,512,0.002630222174856398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,768,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,768,0.011052444577217102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,256,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,512,0.0052177777720822235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,512,0.009736888938479954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,2048,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,256,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,256,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,2048,0.013058666553762225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,64,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,128,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,32,0.0018844445132546956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,128,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,65536,0.025072889195548162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,64,0.005060444275538127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,128,0.00792711145348019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,32,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,51200,0.02051466703414917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,65536,0.046588444047504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,65536,0.054177777634726636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,512,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,16384,0.013365333278973898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,2048,0.005611555443869696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,51200,0.03759466608365377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,51200,0.043656888935301036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,65536,0.10833511087629531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,12288,0.010986666712496014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,16384,0.016182222300105624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,256,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,51200,0.08741866879993015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,12288,0.013145777914259167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,12288,0.01242844429281023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,16384,0.03462666604253981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,8192,0.00904177791542477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,10240,0.011777777638700275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,12288,0.028146667612923518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,8192,0.010176888770527309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,7168,0.008288000192907121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,10240,0.025063999825053748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,8192,0.022280888424979314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,7168,0.009309333231714036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,6144,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,7168,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,7168,0.020265777905782063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,16384,0.016534222496880423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,5120,0.0063991112013657885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,10240,0.008991111483838823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,6144,0.008683555656009251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,5120,0.008628444539176093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,6144,0.01920355525281694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,4096,0.005603555589914322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,5120,0.008343111309740279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,10240,0.011075555450386472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,3584,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,5120,0.01681422193845113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,4096,0.007672000262472365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,4096,0.007606222397751278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,3072,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,3584,0.007479999628331926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,3584,0.007302222152551015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,4096,0.015844444433848064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,2560,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,3072,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,3072,0.0069857777820693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,3584,0.014487110906177096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,2048,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,2560,0.006461333483457565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,2560,0.007061333292060428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,3072,0.013452444639470843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,1536,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,6144,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,2048,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,2048,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,1024,0.0038746665749284956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,2560,0.01351466692156262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,1536,0.006605333338181178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,1536,0.005976888868543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,2048,0.01277599980433782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,768,0.0033111110743549135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,1024,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,1024,0.005398222141795688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,512,0.003165333428316646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,1536,0.011763555308183035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,8192,0.010677333507272931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,768,0.0058986664646201665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,256,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,768,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,512,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,512,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,768,0.010381333529949188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,128,0.0021342221233579847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,256,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,512,0.01039644413524204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,64,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,128,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,256,0.010327111515733931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,32,0.0019395556300878525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,65536,0.019863110449579027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,128,0.0021297778520319196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,128,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,32,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,51200,0.01647644407219357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,65536,0.05407466822200351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,65536,0.045929776297675244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,65536,0.10755022366841634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,51200,0.04279911186960009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,16384,0.0127368892232577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,1024,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,51200,0.0862942205535041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,12288,0.010093332992659675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,16384,0.015090665883488126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,256,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,16384,0.03416711091995239
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,128,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,12288,0.0131217779384719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,10240,0.008346666892369589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,12288,0.012073777616024017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,8192,0.006933333145247565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,10240,0.011728000309732227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,12288,0.027843554814656574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,10240,0.01107377807299296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,7168,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,8192,0.009695111049546136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,10240,0.025253333979182776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,8192,0.0105262224872907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,6144,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,7168,0.01000711073478063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,8192,0.021330666210916307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,6144,0.009155555731720394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,6144,0.008669333325492011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,7168,0.019949333535300363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,5120,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,16384,0.015857777661747403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,4096,0.005927111125654644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,5120,0.008475555313958062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,6144,0.018887110882335238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,5120,0.008061333662933773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,3584,0.005936000082227919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,4096,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,4096,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,5120,0.01683200067943997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,3072,0.0054560001525614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,64,0.00488800017370118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,3584,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,3584,0.007314666277832455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,2560,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,3072,0.00738933351304796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,3072,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,3584,0.015160888433456421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,7168,0.009355555805895064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,3072,0.014104000396198697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,51200,0.03671200076738993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,2560,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,2560,0.006966222491529252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,1536,0.004521777646409141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,2048,0.006597333484225803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,2560,0.013065777719020844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,2048,0.006050666587220297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,1024,0.003479111111826367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,1536,0.005963555640644497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,1536,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,2048,0.012707555459605323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,768,0.003167111012670729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,1024,0.006061333335108227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,1024,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,1536,0.011744000017642975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,512,0.002526222210791376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,768,0.005349333501524395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,1024,0.011397333608733283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,4096,0.015464888678656684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,256,0.0022008888837363985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,512,0.0058782220714622075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,768,0.010670222342014313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,512,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,128,0.002011555557449659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,256,0.005406222409672207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,2048,0.004238222208287981
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,512,0.010374222364690568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,64,0.0018293333964215384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,128,0.004588444613748127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,768,32,0.0018719999740521114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,256,0.01036000003417333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,65536,0.027658666173617046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,64,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,128,0.008351999852392409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,32,0.004570666700601578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,51200,0.021976888179779053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,65536,0.0452479985025194
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,65536,0.05141422152519226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,768,0.005662222289376789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,16384,0.009199110998047723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,65536,0.1075599988301595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,51200,0.03667199942800734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,16384,0.015625778171751235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,16384,0.014666666587193808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,51200,0.08679466777377659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,12288,0.007520000139872233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,256,0.004873777843183941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,10240,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,16384,0.03422044383154975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,128,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,12288,0.012118221984969245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,8192,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,10240,0.010974222587214576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,10240,0.011748444702890186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,7168,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,12288,0.028127110666698877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,8192,0.010392888552612728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,8192,0.009906666974226633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,10240,0.025053333905008104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,6144,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,7168,0.009191999832789103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,7168,0.009924444059530893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,8192,0.021686222818162706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,5120,0.0052871112194326185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,51200,0.038599110311932035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,7168,0.020599999361568026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,4096,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,5120,0.008641777767075432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,5120,0.008014221986134848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,6144,0.019162666466501024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,3584,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,5120,0.016858667135238647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,4096,0.007823111282454597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,12288,0.01310488912794325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,3584,0.00740799970097012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,3072,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,3584,0.006986666884687211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,4096,0.015120888749758402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,2560,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,3072,0.007281777759393056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,3584,0.014464888307783337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,3072,0.006583999842405319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,2048,0.004587555511130227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,2560,0.006115555349323485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,2560,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,3072,0.013787555197874704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,1536,0.003576888806290097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,2048,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,2048,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,6144,0.00903200027015474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,2560,0.013765333427323235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,6144,0.008691555923885768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,1536,0.006324444380071428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,2048,0.012372444073359171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,1536,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,768,0.00253866674999396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,1024,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,4096,0.007620444728268518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,1024,0.006004444426960415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,512,0.002493333278430833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,1536,0.01143200033240848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,1024,0.011338666909270816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,768,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,768,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,256,0.002163555473089218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,512,0.005711110929648082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,512,0.004951111144489712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,128,0.0018320000833935207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,768,0.011352889239788055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,256,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,256,0.004651555584536659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,512,0.009087111386987898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,64,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,32,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,128,0.005547555370463266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,256,0.010750222537252637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,128,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,65536,0.016160888804329764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,64,0.004527111020353106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,128,0.008567999634477828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,32,0.004567111117972268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,65536,0.04385866721471151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,51200,0.03733955489264594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,65536,0.10713066657384236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,51200,0.03452000021934509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,16384,0.00666844430896971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,51200,0.08548533254199558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,12288,0.004927999857399198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,16384,0.01440355512830946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,16384,0.03404888841840956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,12288,0.011923555698659686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,10240,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,12288,0.01308444473478529
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,12288,0.02780088782310486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,51200,0.01402666668097178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,65536,0.046557333734300405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,10240,0.011765333513418833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,10240,0.010936888555685679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,10240,0.025055110454559326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,8192,0.010437332921557956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,7168,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,8192,0.009670221971140968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,8192,0.02162844439347585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,16384,0.015832889411184523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,6144,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,7168,0.008965333302815756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,5120,0.004575110971927643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,6144,0.009322666459613377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,6144,0.008812444077597724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,6144,0.01870933340655433
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,5120,0.008658666577604082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,4096,0.004286222159862518
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,5120,0.00812000036239624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,512,1024,0.0028693332440323303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,4096,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,3584,0.004655111167165968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,4096,0.007646222081449296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,4096,0.015485333071814643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,3584,0.007815111014578078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,3584,0.00665866666369968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,3072,0.004576888763242298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,3584,0.014134221606784396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,2560,0.004628444297446145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,3072,0.006633777585294511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,7168,0.020560888780487906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,3072,0.013756444056828817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,2560,0.006016888966162999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,2560,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,2560,0.013841778039932251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,2048,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,1536,0.0035217776894569397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,8192,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,2048,0.006638222270541721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,2048,0.013082666529549493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,1536,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,1024,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,1536,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,1024,0.005942222144868638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,7168,0.00997688869635264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,1536,0.011359111302428775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,768,0.002566222308410539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,1024,0.011368888947698804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,512,0.0024862223201327855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,768,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,2048,0.004247111164861255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,768,0.00999111102686988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,256,0.0021253333737452826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,512,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,512,0.010399999717871347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,256,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,128,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,256,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,256,0.01033866653839747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,128,0.004538666870858935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,128,0.005612444546487596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,3072,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,1024,0.005623111294375525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,32,0.0018293333964215384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,128,0.00833155545923445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,768,0.005664888769388199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,65536,0.011686222420798408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,64,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,32,0.004951999833186467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,51200,0.010024888647927178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,65536,0.04589688777923584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,65536,0.04145510991414388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,512,0.005616000129116907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,16384,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,51200,0.037343998750050865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,65536,0.10702133178710938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,51200,0.033936000532574125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,5120,0.016813332835833233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,12288,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,16384,0.014098667436175875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,16384,0.015607110328144498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,10240,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,51200,0.08542222446865505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,12288,0.01181333346499337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,12288,0.013078221844302284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,8192,0.004943999979231092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,16384,0.03325422273741828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,256,64,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,12288,0.028098666005664404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,10240,0.011516444385051727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,10240,0.011004444625642566
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,7168,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,8192,0.01034755590889189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,6144,0.00453333349691497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,10240,0.025069332785076563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,7168,0.010043555663691627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,7168,0.0086986662613021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,5120,0.004523555437723796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,6144,0.009368889033794403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,8192,0.021316443880399067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,7168,0.02058844433890449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,6144,0.008674666285514832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,4096,0.004397333496146732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,5120,0.00834222220712238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,6144,0.0188719994492001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,5120,0.00793333351612091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,3584,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,4096,0.0076453329788313965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,4096,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,5120,0.017528888252046373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,3584,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,4096,0.015847111741701763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,3584,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,2560,0.0041884444653987885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,3072,0.006691555596060223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,3072,0.007320889168315464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,3584,0.015122666954994202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,2048,0.0042444442709287005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,2560,0.007254222200976477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,2560,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,3072,0.013811555173661975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,1536,0.003559999995761447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,2048,0.006260444306664997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,2048,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,2560,0.013507555756303998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,1024,0.002895111011134254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,1536,0.006221333311663733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,2048,0.01275644451379776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,768,0.0025173332542181015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,1024,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,1024,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,1536,0.012072000238630505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,512,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,3072,0.004208000169859992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,768,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,1024,0.010494222243626913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,768,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,256,0.001855111163523462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,512,0.005643555687533484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,512,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,128,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,768,0.011042666931947073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,256,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,256,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,512,0.009328000247478485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,64,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,128,32,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,128,0.005537777725193236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,128,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,256,0.009097778134875828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,51200,0.006575999988449945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,64,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,128,0.006955555743641323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,16384,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,12288,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,1536,0.005647111270162795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,10240,0.004566222015354368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,51200,0.03353510962592231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,8192,0.004864000197913912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,16384,0.014488889111412896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,7168,0.004576888763242298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,12288,0.011720000041855706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,10240,0.010405333505736457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,6144,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,8192,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,7168,0.009026666482289633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,5120,0.0038888889054457345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,4096,0.003560888684458203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,6144,0.008336889247099558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,3584,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,8192,0.009644444617960189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,5120,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,3072,0.0041866666740841335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,4096,0.0071937781241205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,2560,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,2048,0.004200889004601372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,3584,0.00665511108107037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,1536,0.003556444413132138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,32,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,3072,0.006664888726340399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,1024,0.0028817777832349143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,768,0.0025271111064487034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,65536,0.0417502224445343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,2560,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,512,0.0024968888610601425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,2048,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,256,0.0020879999630981022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,128,0.0018053334206342697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,1536,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,64,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,1024,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,768,0.005290666802061929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,32,0.0018177777528762817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,512,0.004856889032655292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,256,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,51200,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,64,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,16384,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,32,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,12288,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,65536,0.04140888982348972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,51200,0.033450666401121355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,10240,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,8192,0.003903999924659729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,16384,0.014147554834683737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,7168,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,64,65536,0.006952889263629913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,12288,0.012082666986518435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,6144,0.004515555583768421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,5120,0.004237333519591225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,10240,0.010701333483060202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,4096,0.0038862224254343244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,8192,0.009360888765917884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,3584,0.0041697778635554844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,7168,0.008712000317043727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,6144,0.008344888687133789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,3072,0.003968888686762916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,2560,0.004230222354332606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,5120,0.007927999728255803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,2048,0.004214222232500712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,4096,0.0069982219073507525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,1536,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,3584,0.007033777733643849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,1024,0.0028248888750871024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,3072,0.006616888774765863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,65536,0.0063582224150498705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,768,0.0025235555238193935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,512,0.0021671110557185277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,128,0.004945777770545748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,2048,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,256,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,128,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,1536,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,64,0.0017911110901170308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,1024,0.005196444276306364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,768,0.005316444569163852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,4,32,32,0.0014968889041079415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,512,0.004950222041871813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,256,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,128,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,64,0.004583999928500918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,32,0.005007110950019625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,16384,0.33463467491997617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,12288,0.2551751136779785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,16384,0.22751379013061523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,51200,0.48556264241536456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,10240,0.21306933297051323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,2560,0.005967111223273807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,12288,0.14975466993119982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,12288,0.09947911236021255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,8192,0.17267110612657335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,8192,0.11626133653852676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,10240,0.15271643797556558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,8192,0.0686088866657681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,10240,0.08452977736790974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,7168,0.1421768930223253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,6144,0.12463466326395671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,5120,0.10587110784318711
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,7168,0.09508622354931301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,7168,0.06301510996288724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,6144,0.08609422047932942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,6144,0.053730666637420654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,4096,0.0867982241842482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,16384,0.1354879935582479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,3584,0.07181600067350599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,5120,0.07421955797407362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,4096,0.0655297769440545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,4096,0.03902755512131585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,3072,0.0623920030064053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,51200,1.0787661870320637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,3584,0.037294222248925105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,3584,0.05900177690717909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,3072,0.04684799909591675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,3072,0.035319109757741295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,2560,0.03447733322779337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,2560,0.043030222256978355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,2560,0.0522346662150489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,1536,0.03259199857711792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,51200,0.44745598898993594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,2048,0.028938667641745672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,2048,0.03582488828235202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,1536,0.02316444449954563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,1024,0.022865777214368183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,1536,0.024717332588301763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,768,0.017833777599864535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,512,0.012681778106424542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,1024,0.017458667357762653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,1024,0.02068800065252516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,768,0.015263999501864115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,256,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,5120,0.04623111089070638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,768,0.017279111676745944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,512,0.012851555314328937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,128,0.005132444616821078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,64,0.0041866666740841335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,512,0.013040888640615674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,256,0.010272888673676385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,32,0.0041804446114434134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,128,0.0103511114915212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,256,0.010043555663691627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,128,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,64,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,32,0.008571555217107138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,2048,0.042875554826524526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,16384,0.2545075681474474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,65536,0.5093582471211752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,51200,0.38820799191792804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,51200,0.8532622125413684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,65536,1.086880048116048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,12288,0.18966489368014863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,16384,0.1725848913192749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,16384,0.10715822378794353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,10240,0.1584497822655572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,12288,0.08066488636864556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,51200,0.35350312127007383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,8192,0.12962755892011854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,10240,0.12059199810028076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,10240,0.06845066944758098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,65536,0.45260000228881836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,12288,0.1541244453854031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,7168,0.1136053270763821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,6144,0.09798222117953831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,8192,0.06018400192260742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,7168,0.051458666721979775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,7168,0.07864622275034587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,6144,0.044827553961012095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,8192,0.09246222178141277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,4096,0.07124000125461154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,5120,0.08806044525570339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,5120,0.03897777862019009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,3584,0.06247378057903714
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,5120,0.06056266360812717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,3072,0.0544177790482839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,4096,0.055050667789247304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,2560,0.04548799991607666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,3584,0.03153955605294969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,2048,0.03643200132581923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,3584,0.045801778634389244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,3072,0.03842133283615112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,2560,0.028082665469911363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,1536,0.028239998552534316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,2560,0.03631911012861464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,2048,0.024375999967257183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,1024,0.020043555233213637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,2048,0.03031555480427212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,1536,0.01941244469748603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,768,0.014869333969222175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,1536,0.020801777640978496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,512,0.010372444159454769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,1024,0.015505777464972602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,768,0.013441777891582914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,256,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,512,0.011066666907734342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,512,0.011367999845080905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,128,0.0032097777972618737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,256,0.009379555781682333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,6144,0.06921155585183038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,256,0.00833422193924586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,4096,0.032448000378078885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,128,0.009366222553782994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,32,0.00282222218811512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,128,0.007270221908887227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,64,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,32,0.007713778151406183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,65536,0.3190106550852458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,65536,0.17399288548363578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,51200,0.24829689661661783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,1024,0.01753155555990007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,768,0.014470222923490735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,65536,0.1550648874706692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,16384,0.0839608907699585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,12288,0.06436533398098417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,51200,0.13043555948469374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,64,0.002536888958679305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,16384,0.06718489196565416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,16384,0.042098667886522084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,51200,0.1250071128209432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,10240,0.055436445607079395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,12288,0.033639109796947904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,8192,0.04282755653063456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,12288,0.05155466662512886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,7168,0.03732000125779046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,10240,0.028534223635991413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,6144,0.032187554571363665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,3072,0.029109332296583388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,8192,0.02427644365363651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,8192,0.03685155510902405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,10240,0.04455644554562039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,6144,0.019638222124841478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,6144,0.028102222416136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,5120,0.01722311145729489
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,5120,0.02500711050298479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,4096,0.01514400045077006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,3584,0.020201777418454487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,4096,0.022277333670192297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,3072,0.017515555024147034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,3584,0.015209777487648858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,3584,0.019582221905390423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,3072,0.014808000789748298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,2560,0.014951111541854011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,3072,0.016178665889634024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,2048,0.012496000362767113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,2560,0.01463733282354143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,2560,0.016168889072206285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,1536,0.009706666900051964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,2048,0.014145778285132514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,2048,0.01311288856797748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,1024,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,1536,0.010814222196737925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,7168,0.030543999539481267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,768,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,4096,0.022074666288163926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,1024,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,512,0.0041235557032956015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,768,0.007992888490358988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,1024,0.009319110876984065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,256,0.0025066667132907445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,512,0.007279111279381647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,768,0.007928888830873702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,128,0.0021911110315057966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,512,0.006627555522653792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,256,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,64,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,32,0.002152888932161861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,128,0.006583111153708563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,256,0.0059057776298787855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,128,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,64,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,32,0.005773333211739858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,65536,0.24257956610785592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,51200,0.1871404382917616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,65536,0.15034043788909912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,16384,0.06510755750868055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,5120,0.02693333393997616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,65536,0.136881775326199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,51200,0.1113813320795695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,12288,0.05063022176424662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,1536,0.010791111323568555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,51200,0.1022648877567715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,16384,0.03664888938268026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,12288,0.02955199943648444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,8192,0.03698844379848904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,12288,0.04615555538071526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,7168,0.0324924455748664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,10240,0.025416000021828547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,10240,0.037913779417673744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,8192,0.021282666259341772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,6144,0.028544889556037054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,8192,0.03146133489078946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,7168,0.01960355540116628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,5120,0.024719999896155462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,7168,0.027494221925735474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,6144,0.017847999930381775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,4096,0.020781333247820537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,6144,0.024706666668256123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,5120,0.016121778223249648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,3584,0.018581334087583754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,5120,0.02182577715979682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,4096,0.014451555079884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,3072,0.016506666938463848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,4096,0.020861332615216572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,3584,0.013567999833159976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,2560,0.01423911088042789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,3584,0.017631111873520743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,3072,0.013440888788965015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,2048,0.012238221863905588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,3072,0.014882667197121514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,2560,0.013082666529549493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,1536,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,10240,0.04287199841605293
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,2560,0.014903111590279473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,2048,0.011448889142937131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,16384,0.05579821930991279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,1024,0.006228444476922353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,768,0.005229333208666907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,2048,0.013119999733236102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,1536,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,1024,0.008039111064540016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,512,0.0031555555760860443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,1536,0.010092444717884064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,768,0.0075706665714581805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,256,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,1024,0.008663111262851292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,128,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,768,0.007328000333574083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,512,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,64,0.0018195555441909367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,256,0.006351999938488007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,512,0.006287999865081575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,32,0.0019448887970712448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,256,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,128,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,64,0.005568000177542369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,65536,0.20179466406504312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,32,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,7168,0.021810667382346258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,51200,0.15902311272091335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,65536,0.1403342220518324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,16384,0.05579288800557455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,65536,0.12314132849375407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,12288,0.04323644439379374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,51200,0.10662755701276992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,16384,0.03543377916018168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,10240,0.03646311163902283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,51200,0.09833511379030015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,16384,0.047396444612079196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,8192,0.030812445614073012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,12288,0.02789688772625393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,12288,0.03762311074468825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,10240,0.024266666836208765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,7168,0.027902222341961328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,6144,0.02437688906987508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,8192,0.02439022229777442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,8192,0.02683911058637831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,5120,0.020841777324676514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,7168,0.02273511058754391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,128,0.006629333313968446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,4096,0.018163555198245578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,6144,0.020741333564122517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,5120,0.015493333339691162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,3584,0.015509333875444202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,5120,0.01853155593077342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,4096,0.013813333378897773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,4096,0.01647555496957567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,3584,0.013335111240545908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,3584,0.014987554815080432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,2560,0.0124195557501581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,3072,0.012940444052219391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,3072,0.012961777547995249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,2048,0.01018666641579734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,2560,0.011409777734014722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,1536,0.007616889145639207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,2560,0.012747555143303342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,2048,0.010087110930018956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,1024,0.005929777605666055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,10240,0.03138488862249587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,2048,0.011032000184059143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,1536,0.008786666724416945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,768,0.003843555433882607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,1536,0.008945778012275696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,1024,0.0076657773719893555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,512,0.0032097777972618737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,768,0.007264888948864407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,1024,0.007642666498819987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,6144,0.01814133259985182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,256,0.0021608887861172357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,128,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,768,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,256,0.006272888845867581
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,512,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,128,0.0064071110553211635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,32,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,128,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,64,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,65536,0.16990399360656738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,32,0.005595555735958948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,51200,0.12619288762410483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,65536,0.09223288959927028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,65536,0.07272444168726604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,16384,0.04509600003560384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,7168,0.01907999979125129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,51200,0.06853777832455106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,51200,0.05596622493531969
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,16384,0.022817777262793645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,12288,0.0351128876209259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,16384,0.03559466534190708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,10240,0.028549333413441975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,64,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,12288,0.018581334087583754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,12288,0.02811644474665324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,8192,0.022821333673265245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,10240,0.01627111103799608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,7168,0.02015555567211575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,10240,0.0243848893377516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,8192,0.014490667316648694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,8192,0.02030044462945726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,6144,0.01791111131509145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,7168,0.013400889105266996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,5120,0.015597333510716757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,7168,0.01889511115021176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,6144,0.012739555703269111
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,4096,0.012851555314328937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,6144,0.01649688846535153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,5120,0.011608888705571493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,3584,0.01146133326821857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,5120,0.01479555500878228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,4096,0.010499555203649731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,3072,0.010547555155224271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,4096,0.013738666971524557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,3584,0.009993777506881291
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,2560,0.008973333570692275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,512,0.00674222202764617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,3584,0.012427555190192329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,2048,0.007600888609886169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,3072,0.009773333039548662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,2560,0.008614222208658854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,256,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,1536,0.006206222292449739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,2560,0.010674667027261523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,1024,0.004508444418509801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,2048,0.009676444033781687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,1536,0.007315555380450354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,768,0.0031964443624019623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,1536,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,1024,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,512,0.0028604444944196274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,1024,0.006593777901596493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,768,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,768,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,256,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,512,0.006226666685607698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,128,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,512,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,256,0.005951111101441913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,64,0.0018293333964215384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,256,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,128,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,32,0.0019386667344305252
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,128,0.004862222406599256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,65536,0.14554399914211696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,64,0.0052133335007561584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,32,0.0052693333062860705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,3072,0.014616888430383472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,51200,0.11574399471282959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,2048,0.008115555677149031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,65536,0.08369155724843343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,16384,0.041322665082083807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,12288,0.03198666705025567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,65536,0.06582311126920912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,16384,0.021977777282396953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,51200,0.062112000253465444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,51200,0.053414222266938954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,12288,0.01777777738041348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,16384,0.033497777250077986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,8192,0.023323555787404377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,7168,0.01795377830664317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,10240,0.015650666422314115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,12288,0.02702666653527154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,8192,0.013947554760509066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,10240,0.022308443983395893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,8192,0.020750222934616935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,7168,0.012969777815871768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,5120,0.013408000270525614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,6144,0.011893333660231696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,7168,0.017329777280489605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,4096,0.01201866649919086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,6144,0.01606399979856279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,5120,0.011063999599880643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,3584,0.010332444475756751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,5120,0.014465777410401238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,4096,0.009630222287442949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,3072,0.00925866680012809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,3584,0.008982222113344405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,2560,0.008062221937709385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,3584,0.012067555553383298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,3072,0.010947555303573608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,2048,0.0070186663005087115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,3072,0.010383110907342698
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,2560,0.007265778051482306
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,1536,0.005664000080691443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,2560,0.01034400032626258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,2048,0.007165333463086023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,10240,0.027564444475703772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,1024,0.0038204445607132385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,2048,0.00850755555762185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,1536,0.0069217777086628815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,768,0.003162666741344664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,1536,0.006949333681000604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,1024,0.006331555545330048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,6144,0.015776889191733468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,512,0.002564444517095884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,1024,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,768,0.006133333262470033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,512,0.005974222388532426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,768,0.006242666807439592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,128,0.002128888956374592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,512,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,256,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,64,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,256,0.004997333304749595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,32,0.0017946666727463405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,128,0.005717333406209946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,4096,0.0127688886390792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,128,0.005053333524200651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,3072,0.008430221842394935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,65536,0.12063377433353001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,64,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,51200,0.0960728857252333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,32,0.005201777650250329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,16384,0.03628799981541104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,65536,0.06282755401399401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,65536,0.07324533330069648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,12288,0.028392000330819026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,51200,0.05148355497254265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,10240,0.02404533326625824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,16384,0.020688888099458482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,16384,0.031141334109836157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,12288,0.01720533271630605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,8192,0.020272889071040683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,10240,0.015341333217091031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,256,0.002161777681774563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,10240,0.021168000168270532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,8192,0.019172444939613342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,6144,0.01606044504377577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,7168,0.012645333177513547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,7168,0.016373333003785875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,5120,0.01408266690042284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,6144,0.011755555868148804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,4096,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,6144,0.015174221661355762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,5120,0.010416888528399997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,3584,0.009472000102202097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,5120,0.013916444447305469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,4096,0.008992888861232335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,3072,0.008527999950779809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,4096,0.012488889197508493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,3584,0.008575110799736446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,2560,0.007583111524581909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,3584,0.011043555206722684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,51200,0.05594933032989502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,2048,0.00647288892004225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,3072,0.011190222369299995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,2560,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,1536,0.0053093334039052325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,2560,0.009628444082207149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,12288,0.024951110283533733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,1024,0.0034728890491856467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,2048,0.0069653333889113525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,2048,0.008344000412358178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,768,0.002887111157178879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,1024,0.006360888895061281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,1536,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,512,0.002513777878549364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,1024,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,768,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,256,0.0021404443929592767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,512,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,768,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,128,0.0021013333979580137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,512,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,256,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,64,0.0017955555684036678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,256,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,128,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,32,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,128,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,65536,0.10531910922792222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,3072,0.00777422222826216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,64,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,32,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,51200,0.0838880009121365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,16384,0.031368000639809504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,65536,0.06062044699986776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,65536,0.06666222545835707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,12288,0.025376889440748427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,1536,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,51200,0.050550222396850586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,10240,0.021301334102948506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,16384,0.020023999942673575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,16384,0.02849510974354214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,12288,0.016673778494199116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,8192,0.018760888112915885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,10240,0.014783110883500842
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,12288,0.02329155637158288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,10240,0.0203840004073249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,8192,0.018358222312397428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,8192,0.013059555656380124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,6144,0.014498665928840637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,7168,0.012080888781282636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,7168,0.016199999385409884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,5120,0.012836444709036084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,6144,0.014111111561457316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,5120,0.00965244405799442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,4096,0.010141332944234213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,5120,0.01311822235584259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,7168,0.01814844376511044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,3584,0.008616000413894653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,51200,0.04927466644181145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,4096,0.008365333080291748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,3072,0.007723555796676212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,3584,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,4096,0.011979555918110741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,2560,0.0068986668354935115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,7168,0.01629155543115404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,3072,0.009140444298585257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,3072,0.0074026667409473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,2048,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,2560,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,2560,0.008287111090289222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,1536,0.00434666665063964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,2048,0.007139555282062954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,1024,0.0031662223239739737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,2048,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,1536,0.006599999964237213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,6144,0.011056889262464313
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,1024,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,1024,0.006090666684839461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,512,0.002520888836847411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,768,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,8192,0.01348000019788742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,256,0.0021280000607172647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,512,0.005649777750174205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,512,0.005439110928111606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,128,0.0020231112009949153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,256,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,256,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,64,0.0018213333355055915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,3584,0.010703110860453712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,32,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,128,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,128,0.005021333280536863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,65536,0.0871644417444865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,64,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,32,0.0052888890107472735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,51200,0.06805955701404147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,16384,0.026327111654811438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,65536,0.05865955352783203
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,768,0.0028560000161329904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,51200,0.047297779056761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,12288,0.020055999358495075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,51200,0.04607199960284763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,10240,0.017315554949972365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,16384,0.019461333751678467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,16384,0.027104000250498455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,12288,0.016128000285890367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,8192,0.01499644418557485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,768,0.005713777823580637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,7168,0.013072888884279462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,10240,0.014507555299335055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,8192,0.012717333104875354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,10240,0.018908444378111098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,6144,0.011899555722872416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,7168,0.011729777687125735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,8192,0.01757244434621599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,5120,0.010605333579911126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,7168,0.014462222655614218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,6144,0.00998488896422916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,4096,0.009538667069541084
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,6144,0.013797333670987023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,5120,0.008714666797055138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,3584,0.010662222074137794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,4096,0.008277333445019191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,5120,0.01275644451379776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,3072,0.00905866672595342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,4096,0.01164177805185318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,3584,0.008013333711359236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,2560,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,3584,0.008691555923885768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,2048,0.007624000310897827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,2560,0.0069724445541699724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,3072,0.00851822230550978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,1536,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,1536,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,2560,0.00833777752187517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,1024,0.004289777742491828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,2048,0.007194666398896112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,1536,0.006615110983451207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,768,0.003882666842805015
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,1536,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,12288,0.021743999587164983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,1024,0.006243555496136348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,512,0.0031866667171319327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,1024,0.005967999911970562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,768,0.006136888845099344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,256,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,768,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,512,0.005960000058015187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,512,0.005264889034960005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,128,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,256,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,64,0.0018293333964215384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,128,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,256,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,32,0.0018568889548381169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,128,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,65536,0.07758844561047025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,3072,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,64,0.005114666703674528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,32,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,65536,0.05942310889561971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,51200,0.06157422065734863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,16384,0.023205333285861548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,65536,0.05785777833726671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,65536,0.056594669818878174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,12288,0.018000889155599806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,2048,0.0069120000633928515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,51200,0.04641600118743049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,16384,0.019268444842762418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,51200,0.044138666656282216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,16384,0.0236835562520557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,8192,0.013351110948456658
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,12288,0.020039111375808716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,7168,0.012112889024946423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,8192,0.0123848890264829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,10240,0.018079110317760043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,6144,0.011046222514576383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,8192,0.015119999647140503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,7168,0.011394666300879585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,5120,0.009813333551088968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,7168,0.013771555489963956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,6144,0.00979288915793101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,4096,0.008966222405433655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,6144,0.013122667041089801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,5120,0.009010666774378883
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,3584,0.009576888548003303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,5120,0.012062222593360476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,4096,0.008286221987671321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,3072,0.008946666287051307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,4096,0.009365333451165093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,3584,0.007991111112965478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,3584,0.008008889026112026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,3072,0.007592889169851939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,2048,0.006558222075303395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,10240,0.015526221858130561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,2560,0.0069520001610120135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,12288,0.015896000795894198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,1536,0.005196444276306364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,2560,0.007996444072988298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,2048,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,1024,0.0038924444880750445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,10240,0.014342221948835584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,2048,0.006608888920810487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,1536,0.006604444649484422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,768,0.0033822223130199644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,512,0.00282577777074443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,1024,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,1536,0.006076444354322221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,768,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,1024,0.005970666805903117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,768,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,512,0.005884444548024072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,128,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,512,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,256,0.005566222386227713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,64,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,32,0.0018595556418100994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,256,0.005154666801293691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,128,0.005550222264395819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,65536,0.06695644723044501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,128,0.004863111095296012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,64,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,51200,0.05275999837451511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,32,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,2560,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,65536,0.0541928874121772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,16384,0.019700444406933255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,65536,0.05667377842797173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,12288,0.015816888875431485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,3072,0.008640000389681922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,51200,0.04587555593914456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,10240,0.013904000322024027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,16384,0.018848000301255118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,51200,0.04253155655331082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,8192,0.012223111258612739
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,16384,0.022302221920755174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,12288,0.01568000018596649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,7168,0.011092444260915121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,12288,0.0196079992585712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,6144,0.010384000009960597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,8192,0.011611555185582904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,10240,0.01660977800687154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,7168,0.010390222072601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,8192,0.015170666906568738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,5120,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,256,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,6144,0.009422221945391761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,4096,0.008043555749787224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,6144,0.012366222010718452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,5120,0.009009777671760982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,3584,0.007456888755162557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,4096,0.008258666429254744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,5120,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,3072,0.008266666697131263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,4096,0.008839999636014303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,3584,0.007675555845101674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,2560,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,3584,0.00813333359029558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,2048,0.006217777729034424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,3072,0.007270221908887227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,2560,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,1536,0.005199999858935674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,2560,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,2048,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,1024,0.0042204442951414315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,2048,0.0069351109365622205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,1536,0.006586666736337874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,768,0.004107555581463708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,1536,0.005932444499598608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,10240,0.013798221945762634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,1024,0.006229333165619109
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,512,0.003162666741344664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,768,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,1024,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,256,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,768,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,512,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,128,0.002222222172551685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,7168,0.013281777501106262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,256,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,64,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,512,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,32,0.0018462222069501877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,256,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,128,0.005658666706747479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,65536,0.05903377797868517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,128,0.004927111168702443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,64,0.004934222333961063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,65536,0.05584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,32,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,3072,0.007662222617202335
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,16384,0.018059555027219985
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,65536,0.051797333690855235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,12288,0.014224888549910652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,51200,0.045187556081347995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,10240,0.012368888490729861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,51200,0.04105866617626614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,16384,0.018684445155991446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,8192,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,12288,0.01516177753607432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,7168,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,16384,0.02119111186928219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,10240,0.013071111506885953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,12288,0.017736888594097562
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,6144,0.009018667042255402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,8192,0.010710222025712332
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,10240,0.014099554883109199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,5120,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,7168,0.010141332944234213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,4096,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,6144,0.009454222189055549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,7168,0.012718222207493253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,3584,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,5120,0.00887288898229599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,6144,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,3072,0.007811555431948767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,4096,0.008196444147162968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,3584,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,4096,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,2560,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,3072,0.007465778125656976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,2048,0.005912888795137405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,3072,0.007118222614129384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,2560,0.00720977783203125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,51200,0.045744889312320285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,1536,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,2560,0.006647999915811751
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,2048,0.0069155556460221606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,1024,0.004113777644104428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,1536,0.006642666541867786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,2048,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,768,0.0035822221802340615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,1024,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,1536,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,512,0.003132444495956103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,1024,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,768,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,256,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,768,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,512,0.005624889085690181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,128,0.0021662221600612006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,8192,0.013850666582584381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,256,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,512,0.005225777626037598
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,64,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,32,0.0018684443914228016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,256,0.005033777819739448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,128,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,65536,0.048306667142444186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,5120,0.008655110994974772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,128,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,51200,0.0362462235821618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,64,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,32,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,16384,0.014112000664075216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,65536,0.05507377783457438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,65536,0.049547556373808116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,3584,0.007329777710967594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,12288,0.01149511088927587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,51200,0.0445493327246772
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,10240,0.01036444471942054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,51200,0.03990933299064636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,12288,0.014186667071448432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,8192,0.009463999834325578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,16384,0.02049688829316033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,10240,0.012698666916953193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,7168,0.009022222624884712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,12288,0.015402666396564908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,6144,0.007976888782448238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,8192,0.010984888507260216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,7168,0.010088889135254754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,8192,0.013056889176368713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,5120,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,6144,0.009496889180607265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,7168,0.0107351111041175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,4096,0.008972444468074376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,6144,0.009000889129108852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,3584,0.007948444121413762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,5120,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,4096,0.00797155582242542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,3072,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,4096,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,3584,0.007633777956167857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,3584,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,3072,0.007310222420427535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,3072,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,2048,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,2560,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,2560,0.006592888798978593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,16384,0.018216888109842937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,2048,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,2048,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,1024,0.0041662222809261745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,10240,0.014094221923086377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,1536,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,768,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,1024,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,1024,0.005532444351249271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,5120,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,512,0.0031902222997612427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,768,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,256,0.0024177777684397167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,768,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,512,0.005579555614127054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,128,0.0021804444905784395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,256,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,512,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,64,0.0019022222194406723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,256,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,32,0.0018275556051068837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,128,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,65536,0.03595022360483805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,128,0.004866666677925321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,51200,0.029081778393851385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,1536,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,32,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,65536,0.05473866727617052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,16384,0.012392889294359418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,65536,0.04834666517045763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,51200,0.044087111949920654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,1536,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,12288,0.009670221971140968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,16384,0.017320000463061862
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,51200,0.03852799865934584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,16384,0.017824888229370117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,12288,0.01313244468635983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,8192,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,10240,0.01204355557759603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,7168,0.00795022232664956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,10240,0.01237511138121287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,8192,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,6144,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,8192,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,7168,0.010043555663691627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,5120,0.006225777582989798
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,2560,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,6144,0.009328888522254096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,4096,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,6144,0.008797333472304875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,5120,0.008654221892356873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,3584,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,5120,0.008223999705579545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,4096,0.00794577764140235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,64,0.004928888960017098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,4096,0.007344000041484833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,3072,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,3584,0.007641777396202087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,2560,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,3584,0.007248000138335758
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,3072,0.007685333490371704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,2048,0.005504888792832692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,10240,0.008799110849698385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,3072,0.006969778074158563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,2560,0.006991110742092133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,1536,0.004828444371620814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,2560,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,2048,0.0069395556218094295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,1024,0.003847111016511917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,12288,0.014815110299322339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,1536,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,2048,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,768,0.0035413333939181436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,512,0.0028453332682450614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,1536,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,1024,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,256,0.0021724444296624926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,768,0.005955555372767978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,1024,0.005504888792832692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,512,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,128,0.0021271111650599376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,768,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,7168,0.009350222018029954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,64,0.0018026667336622875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,512,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,256,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,32,0.0018115554832749898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,65536,0.025215999947653875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,128,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,256,0.005036444299750858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,51200,0.020604444874657523
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,128,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,16384,0.013255111045307584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,65536,0.05416533350944519
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,64,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,65536,0.04647022154596117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,51200,0.043848001294665866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,51200,0.037308444579442344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,10240,0.009101333717505137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,16384,0.015843555331230164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,16384,0.016182222300105624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,12288,0.01311644415060679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,8192,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,12288,0.012150222228633033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,10240,0.01179377817445331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,7168,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,10240,0.011021333436171213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,8192,0.010713777608341642
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,6144,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,7168,0.01000088867213991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,8192,0.009778666827413771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,5120,0.00664533343580034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,7168,0.009347555538018545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,4096,0.00591111100382275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,6144,0.009296889106432596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,3584,0.0053093334039052325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,6144,0.008676444490750631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,5120,0.008623110751310984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,4096,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,3072,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,5120,0.008072000410821702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,3584,0.007304888632562425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,3072,0.007262222468852997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,3072,0.00683111117945777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,2048,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,32,0.004967111266321606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,12288,0.010678222609890832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,1536,0.005899555567238066
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,2048,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,2048,0.006262222097979651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,1024,0.0036115555299652945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,1536,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,1536,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,768,0.0038986665507157645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,1024,0.006291555447710886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,1024,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,512,0.0028888889484935333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,768,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,768,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,256,0.0024666666156715816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,512,0.0057813334796163775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,2560,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,512,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,128,0.002132444539003902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,3584,0.007250666618347168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,256,0.005580444302823808
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,64,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,32,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,256,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,128,0.005214222189452913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,65536,0.019840000404251944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,2560,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,51200,0.01641688909795549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,64,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,32,0.004876444323195351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,65536,0.05406755540106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,16384,0.013464888764752282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,65536,0.04554399847984314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,51200,0.043523556656307645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,51200,0.0369484457704756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,12288,0.009047110875447592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,16384,0.01575555569595761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,16384,0.015117333994971381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,12288,0.01310400002532535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,12288,0.01223111069864697
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,8192,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,10240,0.011925333076053195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,10240,0.010716444088353051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,4096,0.0074035558435651995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,7168,0.007088888850477006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,8192,0.009687999884287516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,6144,0.006272000157170826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,7168,0.00998755544424057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,2560,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,5120,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,7168,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,6144,0.009302222066455418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,4096,0.005882666756709416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,6144,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,5120,0.008541333178679148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,3584,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,5120,0.00795555528667238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,4096,0.007640000018808577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,3072,0.0053182223604785065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,4096,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,3584,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,2560,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,3584,0.007237333390447829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,10240,0.008345777789751688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,2048,0.0042151109211974675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,3072,0.006618666566080517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,2560,0.0069475554757648045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,1536,0.004865777575307422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,2560,0.006264888991912206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,2048,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,1024,0.0035546666218174826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,2048,0.005940444353553984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,8192,0.010666666759385003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,768,0.0032186667538351486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,1536,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,1536,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,1024,0.005910222315125995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,512,0.0028737777223189673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,768,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,1024,0.005372444374693765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,256,0.0021537778278191886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,768,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,512,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,128,0.0021235555824306277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,256,0.00554577757914861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,512,0.0052435555391841466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,64,0.0017973333597183228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,768,32,0.001814222170246972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,128,0.005527110977305307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,65536,0.02732266651259528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,128,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,51200,0.022658665974934895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,64,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,32,0.004595555365085602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,65536,0.05099022057321337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,16384,0.008953777452309927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,3072,0.007247111035717859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,12288,0.007313778003056844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,51200,0.03840000099605984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,51200,0.036406222316953875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,16384,0.015821332732836407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,10240,0.0069466663731469055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,16384,0.014471999473041959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,12288,0.013047999805874295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,8192,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,10240,0.011741333537631564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,12288,0.012090666426552666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,7168,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,10240,0.011025778121418424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,8192,0.010703110860453712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,6144,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,8192,0.009661333428488838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,128,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,5120,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,7168,0.009359111388524374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,6144,0.009309333231714036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,4096,0.005592888842026393
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,6144,0.00870222184393141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,5120,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,3584,0.005597333527273602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,256,0.004913777940803104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,5120,0.007972444097201029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,4096,0.007986666427718269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,3072,0.005040888984998067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,2560,0.004922666483455234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,4096,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,3584,0.007675555845101674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,2048,0.00462400002612008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,3072,0.007301333049933116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,3584,0.006991110742092133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,2560,0.006882666713661618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,3072,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,65536,0.04510044389300876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,2560,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,2048,0.006575111299753189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,1024,0.0029084444459941653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,2048,0.006120000034570694
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,1536,0.006498666687144174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,1024,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,1536,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,512,0.0024542222834295696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,1024,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,768,0.005987555616431766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,768,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,256,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,7168,0.009696889254781935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,128,0.0021422221842739317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,512,0.005437333136796951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,256,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,64,0.0018133332745896445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,256,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,128,0.005565333283609814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,32,0.0017928888814316855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,65536,0.01622488929165734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,128,0.004580444345871608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,64,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,51200,0.014466666513019137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,65536,0.04587999979654948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,32,0.004889777965015835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,16384,0.006591111007663939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,1536,0.0035208890007601846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,65536,0.04433866673045688
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,12288,0.005208888815508948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,51200,0.03719733489884271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,51200,0.034232000509897866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,512,768,0.002838222309947014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,10240,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,16384,0.014434667097197639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,12288,0.01309333327743742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,8192,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,12288,0.012085333466529846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,10240,0.011752888560295105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,7168,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,8192,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,10240,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,6144,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,512,0.005341333233647876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,7168,0.00980711148844825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,5120,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,7168,0.009285333255926767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,6144,0.009363555245929295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,4096,0.004567999806669024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,5120,0.008616000413894653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,6144,0.00870222184393141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,3584,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,5120,0.008039111064540016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,4096,0.007681777907742395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,3072,0.004552000098758274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,4096,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,3584,0.007602666815121968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,2560,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,3072,0.007528000407748752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,3584,0.0069439998931354946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,2048,0.004521777646409141
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,3072,0.006558222075303395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,2560,0.006963555183675554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,1536,0.0035253332720862497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,2048,0.006897777732875612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,2560,0.0063182223174307085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,16384,0.015816888875431485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,2048,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,1536,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,768,0.002509333400262727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,1536,0.006142222219043308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,512,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,1024,0.0052284445199701525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,768,0.005934222290913264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,256,0.0021742222209771476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,768,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,8192,0.009725333087974125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,512,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,512,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,256,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,32,0.0018168888572189545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,128,0.0052471111218134565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,256,0.004968889057636261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,65536,0.01107377807299296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,128,0.004567999806669024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,51200,0.009656888743241629
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,64,0.004600000050332812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,32,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,16384,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,65536,0.0417795545525021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,51200,0.037008888191647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,51200,0.033442666133244835
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,12288,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,16384,0.015809777710172866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,10240,0.005239999956554837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,16384,0.014446222119861178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,12288,0.013058666553762225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,1024,0.005973333285914526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,12288,0.011744000017642975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,8192,0.005201777650250329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,10240,0.011704000333944956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,7168,0.004920000003443824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,10240,0.011023999916182624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,8192,0.010403555300500657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,128,0.0018666666001081467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,6144,0.004871110949251386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,8192,0.009689778089523315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,7168,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,5120,0.004260444392760594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,6144,0.009325332939624786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,7168,0.009144888983832465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,4096,0.004361777669853634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,5120,0.00846400029129452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,6144,0.00867022242810991
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,3584,0.004579555657174853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,5120,0.00794400026400884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,4096,0.007779555188284979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,3072,0.004545777622196409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,65536,0.045844445625940956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,3584,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,4096,0.007243555453088548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,3584,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,3072,0.007287110719415877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,2048,0.004243555582231945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,2560,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,3072,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,1536,0.0031991110493739447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,256,1024,0.002858666703104973
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,2048,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,2560,0.006324444380071428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,2048,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,1536,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,768,0.00254488881263468
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,1024,0.005959110955397288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,1536,0.005678222411208683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,512,0.002524444419476721
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,1024,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,256,0.0019911111642917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,512,0.005609777652555042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,768,0.00498577786816491
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,512,0.005108444641033809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,256,0.0052800000541739995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,64,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,256,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,32,0.0018035556293196147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,128,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,65536,0.006619555668698416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,51200,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,64,0.004869333157936732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,2560,0.004524444540341695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,32,0.004902222090297275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,12288,0.00508355556262864
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,65536,0.04143288731575012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,10240,0.004541333350870344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,51200,0.03360355562633938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,8192,0.004906666775544484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,16384,0.014088888963063559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,7168,0.004595555365085602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,12288,0.012073777616024017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,10240,0.010689777632554373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,8192,0.009659555223253038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,5120,0.004361777669853634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,7168,0.008975110948085785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,4096,0.003911111089918348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,6144,0.008442666795518663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,128,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,3584,0.004513777792453766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,3072,0.0041697778635554844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,5120,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,2560,0.004536889079544279
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,4096,0.0069751110341813825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,3584,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,128,0.004535111288229625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,1536,0.0032666667054096856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,3072,0.0062977779242727495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,1024,0.0028239999794297745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,16384,0.005155555489990446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,768,0.002518222149875429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,2048,0.005892444401979446
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,512,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,1536,0.005610666755172942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,256,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,1024,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,128,1024,0.0028577778074476454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,128,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,64,0.0015075555485155846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,6144,0.004582222137186262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,512,0.004638222356637319
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,32,0.0018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,65536,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,256,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,128,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,51200,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,64,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,16384,0.004965333475006951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,32,0.004552000098758274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,12288,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,65536,0.04177600145339966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,10240,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,51200,0.033556444777382746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,8192,0.004208000169859992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,16384,0.014085332552591959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,7168,0.004063999901215236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,12288,0.011752888560295105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,6144,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,2560,0.00629333323902554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,10240,0.011030221978823343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,5120,0.0042239998777707415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,8192,0.009633777870072259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,7168,0.009009777671760982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,3584,0.0041928887367248535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,6144,0.00832088871134652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,5120,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,768,0.005426666802830166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,4096,0.007270221908887227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,3584,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,2048,0.0038888889054457345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,3072,0.006800888727108638
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,1536,0.003539555602603488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,2048,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,1024,0.0028026666906144884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,1536,0.005923555543025334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,768,0.0025128889828920364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,64,2048,0.004173333446184794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,512,0.002135111019015312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,1024,0.00519466648499171
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,4096,0.0041928887367248535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,256,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,768,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,128,0.001777777862217691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,64,0.0016177777821818988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,512,0.004583999928500918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,32,0.001642666653626495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,3072,0.0039022221333450745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,3,32,2560,0.004234666625658671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,128,0.004893333133724001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,256,0.004959110998445087
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,64,0.004513777792453766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,32,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,2560,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,16384,0.3336302174462213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,12288,0.25319022602505153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,16384,0.2408533361223009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,51200,0.5274817678663466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,10240,0.21213600370619032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,12288,0.1510888867908054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,12288,0.10674044158723618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,8192,0.17287288771735299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,10240,0.08739288647969563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,8192,0.11785954899258083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,8192,0.06992799705929227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,7168,0.14104089472028944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,6144,0.12307111422220866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,7168,0.09637688928180271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,51200,0.3984791172875299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,7168,0.06448088751898871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,5120,0.10443288750118679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,6144,0.06088266770044962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,4096,0.08666221963034736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,6144,0.10486755768458049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,5120,0.0550871127181583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,16384,0.13517599635654026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,3584,0.07149866554472181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,3072,0.06163022253248426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,4096,0.06807911396026611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,4096,0.04900088906288147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,3584,0.045610666275024414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,3584,0.05364711085955302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,2560,0.05229066809018453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,2048,0.04219288958443535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,3072,0.03987288806173537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,3072,0.04866400029924181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,2560,0.03350311186578538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,1536,0.03180088930659824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,2560,0.04049333267741733
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,1024,0.022409778502252366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,2048,0.029105779197480943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,2048,0.036426666710111834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,1536,0.023559111687872145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,1024,0.018000889155599806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,1536,0.028135998381508723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,512,0.012657778130637275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,1024,0.020938666330443487
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,768,0.01538755496342977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,256,0.008002666963471307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,512,0.012826667063766055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,5120,0.07296533054775663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,128,0.004966222163703707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,256,0.010394666757848527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,256,0.010033778018421596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,64,0.004202666795916027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,128,0.010390222072601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,10240,0.1284631093343099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,32,0.004201777693298128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,64,0.008338666624493068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,32,0.008210666477680206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,768,0.01759111053413815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,768,0.017148445049921673
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,512,0.012992888689041138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,65536,0.5126906500922309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,51200,1.1215857399834526
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,65536,1.0873182084825304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,16384,0.2521724436018202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,51200,0.8481493526034884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,51200,0.460935115814209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,12288,0.1881128946940104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,16384,0.15198755264282227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,12288,0.12196444140540229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,16384,0.11110488573710124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,51200,0.3197573290930854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,10240,0.1561973359849718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,65536,0.43314488728841144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,128,0.008093333078755273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,8192,0.128400895330641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,10240,0.10236177841822307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,7168,0.11188177267710368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,10240,0.07079822487301297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,8192,0.05877421961890327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,6144,0.09710488716761272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,8192,0.09626489215426975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,7168,0.07909066809548272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,7168,0.05239555570814344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,5120,0.08902133173412746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,6144,0.07380444473690458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,4096,0.07162400086720784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,6144,0.04925600025388929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,3584,0.062183108594682485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,5120,0.06016977628072103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,3072,0.05448711249563429
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,4096,0.03868088788456387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,4096,0.05673155519697401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,2560,0.04546044601334465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,3584,0.03772799836264716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,3584,0.04497155547142029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,2048,0.036842667394214206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,3072,0.03325600094265408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,3072,0.041449778609805636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,2560,0.032595554987589516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,2048,0.024319999747806128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,1024,0.020244444410006206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,2048,0.03067644437154134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,1536,0.019928889142142404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,1536,0.02295555505487654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,768,0.014826666977670459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,1024,0.017529777354664274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,12288,0.083569778336419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,512,0.010409778190983666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,768,0.013582222163677216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,256,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,512,0.011401777466138205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,5120,0.04361866580115425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,512,0.011387555963463254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,128,0.0031519999934567344
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,256,0.009351111120647853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,64,0.00258933338854048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,256,0.008310221963458592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,32,0.0026026666164398193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,128,0.009386666946940953
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,128,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,1536,0.028128001425001357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,64,0.007870222131411234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,32,0.007261333366235097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,65536,0.31615376472473145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,51200,0.24545955657958984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,65536,0.17351110776265463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,1024,0.015296889675988091
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,16384,0.08338399728139241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,65536,0.15598666667938232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,51200,0.1706622176700168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,768,0.014204444156752693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,16384,0.04236888885498047
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,10240,0.05479200018776787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,51200,0.11838044060601129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,16384,0.06804177496168348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,8192,0.0436062216758728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,12288,0.033153778976864286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,12288,0.04960888955328199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,7168,0.038395557138654925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,10240,0.028775112496482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,8192,0.02520622147454156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,8192,0.03733777668741014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,6144,0.033207999335394964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,5120,0.028130667077170476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,7168,0.022856000396940444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,10240,0.043839110268486865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,7168,0.03143822153409322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,4096,0.023031999667485554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,5120,0.020992888344658747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,3584,0.020592888196309406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,4096,0.018823999497625563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,6144,0.029672887590196397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,3072,0.01798488861984677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,3584,0.018204443984561496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,3584,0.019606222709019978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,4096,0.022152000003390845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,2560,0.014751111467679342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,3072,0.016601777738995023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,3072,0.01812177730931176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,2048,0.013016000390052795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,2560,0.014826666977670459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,2560,0.013792888985739814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,1536,0.010279110736317104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,2048,0.012907555533779992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,2048,0.014187556174066333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,2560,0.027719999353090923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,1024,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,1536,0.010791999598344168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,6144,0.02200088898340861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,768,0.005961777849329843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,1536,0.011749332977665795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,1024,0.009156444834338294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,512,0.004283555679851108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,5120,0.025078222155570984
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,1024,0.009398221969604492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,256,0.0024826667375034755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,512,0.007321777443091075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,768,0.008045333127180735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,128,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,512,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,256,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,64,0.0021111110432280437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,128,0.00674044465025266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,32,0.0021297778520319196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,12288,0.0637342201338874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,128,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,64,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,65536,0.24135377671983507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,32,0.005557333429654439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,65536,0.14657421906789145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,51200,0.18515200085110137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,16384,0.0638399985101488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,65536,0.1336008840137058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,12288,0.04920622375276354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,51200,0.10164444314108954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,51200,0.14926932917700872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,16384,0.05573066737916735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,10240,0.042077332735061646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,16384,0.0377422206931644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,8192,0.03713511096106635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,768,0.00796888851457172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,12288,0.03506311111980014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,12288,0.042632887760798134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,7168,0.03243466549449497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,10240,0.025576889514923096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,6144,0.02832088867823283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,8192,0.021819555097156104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,10240,0.03856888744566176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,5120,0.024736889534526404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,8192,0.031239999665154353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,256,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,4096,0.020875554945733812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,6144,0.01888177792231242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,3584,0.018566222654448617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,6144,0.02685244381427765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,5120,0.0185217774576611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,5120,0.021235555410385132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,3072,0.016562667157914903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,4096,0.02051466703414917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,3584,0.016415110892719693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,2560,0.014430221584108142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,3072,0.014833778142929077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,2048,0.012086222569147745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,3072,0.016140444411171805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,2560,0.012986666626400419
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,1536,0.008237332933478886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,2560,0.013083555632167392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,2048,0.011128889189826118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,1024,0.006281777802440856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,2048,0.0131217779384719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,1536,0.009670221971140968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,768,0.005075555708673266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,1536,0.010983111129866706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,1024,0.008356444537639618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,512,0.0032417778339650896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,1024,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,768,0.007631999750932057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,256,0.002528000002106031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,7168,0.02034133341577318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,768,0.007325333025720384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,512,0.006984000404675801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,128,0.0021484444538752237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,7168,0.027802666028340656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,512,0.006312888943486744
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,64,0.0020088888704776764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,256,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,32,0.0021768889079491296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,128,0.006365333166387346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,256,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,128,0.005320888840489917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,65536,0.201218671268887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,64,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,3584,0.017532444662517972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,32,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,51200,0.1582497755686442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,16384,0.05545333358976576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,65536,0.13984444406297472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,65536,0.12167199452718098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,12288,0.042472889026006065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,16384,0.04833333359824287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,51200,0.09642577833599514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,51200,0.12680799431271023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,10240,0.036419557200537786
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,8192,0.030232889784706965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,12288,0.03763822383350796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,12288,0.027867555618286133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,10240,0.02607733342382643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,7168,0.02775466607676612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,8192,0.02054222259256575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,6144,0.024400000770886738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,7168,0.01940000057220459
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,7168,0.023723555935753718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,4096,0.01687111126052009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,5120,0.020724443925751578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,6144,0.01834844383928511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,6144,0.021353777911927965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,5120,0.017092444830470614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,5120,0.017993777990341187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,3584,0.01591022147072686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,4096,0.015275556180212231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,4096,0.01681422193845113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,3072,0.014885332849290637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,16384,0.035531553957197405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,3584,0.015265777707099915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,2560,0.012758221891191272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,3072,0.01311466677321328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,3072,0.013603555659453074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,2048,0.009691555466916826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,10240,0.031143112315071955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,2560,0.011891555454995898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,2560,0.011863999896579318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,1536,0.007928888830873702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,8192,0.02682133350107405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,2048,0.010225777824719747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,1024,0.006042666733264923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,2048,0.011338666909270816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,1536,0.008936888641781276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,768,0.003937777959638172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,1536,0.009399111072222391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,512,0.0032053333189752367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,4096,0.018423111902342904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,768,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,1024,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,256,0.002480000050531493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,768,0.006928000185224745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,512,0.006595555692911148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,256,0.006267555471923616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,512,0.006325333482689328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,256,0.005320000151793162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,32,0.002013333348764314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,128,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,3584,0.014399111270904541
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,128,0.004953777624501122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,65536,0.1653991142908732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,64,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,51200,0.12586577733357748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,32,0.005286222116814719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,16384,0.04454666707250807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,65536,0.07101688782374065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,12288,0.035154667165544294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,51200,0.05680266353819105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,51200,0.0892382197909885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,16384,0.03494577937655979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,16384,0.023026666707462732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,1024,0.00795111142926746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,12288,0.028767999675538804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,8192,0.0236791107389662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,10240,0.024367999699380662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,7168,0.021939555803934734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,128,0.002130666747689247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,8192,0.014512888259357877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,8192,0.02091022166940901
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,7168,0.013751111096805997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,7168,0.018333334061834548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,5120,0.016195555528004963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,6144,0.013061333033773633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,6144,0.017844445175594754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,4096,0.013771555489963956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,5120,0.012615111139085559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,65536,0.08952089150746663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,5120,0.014494222071435718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,3584,0.012060444388124677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,4096,0.011318222516112857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,4096,0.01386666629049513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,3072,0.010456889039940305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,10240,0.029142220815022785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,3584,0.010818666881985135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,2560,0.008998221821255153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,3584,0.012091555529170565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,3072,0.009826666778988307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,12288,0.018590221802393597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,2048,0.007990222010347579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,1536,0.00665866666369968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,10240,0.016433777080641854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,3072,0.011235555840863122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,2048,0.008362666600280339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,1024,0.004577777865860197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,2560,0.01035644445154402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,2048,0.009838222629494136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,768,0.0032951111594835916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,6144,0.018374222848150466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,1536,0.007674666742483775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,512,0.0028595555987623003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,1024,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,1024,0.006620444357395172
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,256,0.0022808888720141514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,512,0.005992888990375731
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,768,0.006309333360857434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,128,0.002136888810329967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,512,0.005656888915432825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,256,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,64,0.0020026666008763844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,256,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,32,0.0018764444523387486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,128,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,64,0.0052044445441828836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,65536,0.1442977719836765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,32,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,2560,0.009006222089131674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,51200,0.11531910631391738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,1536,0.00739822205570009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,65536,0.08295644654168023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,65536,0.06530133220884535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,16384,0.04105688797103034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,12288,0.03222666515244378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,51200,0.08344622453053792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,51200,0.0531582236289978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,768,0.006355555521117316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,10240,0.027247111002604168
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,16384,0.021912889348136053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,8192,0.023261333505312603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,16384,0.03394577900568644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,12288,0.017825777331988018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,7168,0.018560000591807894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,12288,0.027344889110989038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,10240,0.015617777903874716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,6144,0.015953777564896476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,8192,0.0138586668504609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,10240,0.0236471113231447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,5120,0.013814222481515674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,7168,0.012888000243239932
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,7168,0.017867555220921833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,6144,0.01205599970287747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,4096,0.012311110893885294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,6144,0.016574222180578444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,5120,0.011036444041464063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,3584,0.010434666441546546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,5120,0.015444444285498725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,4096,0.009908444351620143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,3072,0.009491555392742157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,3584,0.009000889129108852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,2560,0.008283555507659912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,3584,0.012279111478063794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,3072,0.008391111261314815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,2048,0.007059555914666917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,3072,0.011146666275130378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,2560,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,1536,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,2560,0.009335111412737105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,1024,0.0039022221333450745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,2048,0.007341333561473423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,1536,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,2048,0.008968888885445064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,768,0.0032097777972618737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,1024,0.006311111152172089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,1536,0.007354666789372762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,512,0.0028364445186323593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,1024,0.006294222341643439
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,768,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,256,0.002165333264403873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,512,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,8192,0.020562666985723708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,128,0.0021600000974204806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,512,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,64,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,256,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,128,0.005622222191757626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,32,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,128,0.005112888912359874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,65536,0.12444355752733018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,4096,0.013064888616402945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,64,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,32,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,51200,0.09565777911080255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,16384,0.036066667901145086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,65536,0.07307822174496122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,65536,0.06303999821345012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,12288,0.02752444479200575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,51200,0.05115200082461039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,51200,0.07591466771231757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,10240,0.023984889189402264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,12288,0.01717333330048455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,16384,0.032695998748143516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,8192,0.020215999748971727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,12288,0.02529688841766781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,128,0.005938666562239329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,10240,0.015152000718646579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,10240,0.02159999973244137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,768,0.006075555665625467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,7168,0.018177777528762817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,6144,0.0158933334880405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,256,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,8192,0.013464888764752282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,5120,0.014112888111008538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,8192,0.018746667438083224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,7168,0.012718222207493253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,4096,0.010860444770918952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,6144,0.011695110963450538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,7168,0.016951999730534024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,5120,0.010495111346244812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,6144,0.015839111473825242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,3584,0.009510222408506606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,3072,0.00848711116446389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,5120,0.01349511080318027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,4096,0.012702222499582501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,3584,0.008936000367005667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,2560,0.0076657773719893555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,3072,0.008362666600280339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,3584,0.011238222320874533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,2048,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,3072,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,2560,0.007629333270920648
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,1536,0.0053546664615472155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,2560,0.008770667016506195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,16384,0.02075377768940396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,2048,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,1024,0.0036862223512596558
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,768,0.0031866667171319327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,2048,0.00795733349190818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,1536,0.006656000183688269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,512,0.0025964443468385274
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,1024,0.006286222073766921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,1536,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,768,0.0059573331640826324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,256,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,1024,0.006315555423498154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,128,0.0020151111400789684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,768,0.005991111199061076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,512,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,256,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,64,0.001810666587617662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,32,0.0018159999615616268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,128,0.00573244442542394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,256,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,128,0.004993777722120285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,65536,0.10422755612267388
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,64,0.005205333232879639
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,51200,0.08238755332099067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,4096,0.009025777379671732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,32,0.005063111169470681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,16384,0.030939555830425684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,65536,0.06722578075197008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,12288,0.02516533268822564
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,65536,0.06061244673199124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,16384,0.01999022232161628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,51200,0.049010667535993785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,51200,0.07316622469160292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,16384,0.02875822120242649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,12288,0.016509332590632968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,8192,0.018382221460342407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,12288,0.023744000328911677
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,7168,0.01604622271325853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,10240,0.01479111115137736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,8192,0.013226666384273104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,10240,0.021474666065639917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,7168,0.012132444315486483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,8192,0.018889778190188937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,5120,0.012749333348539142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,7168,0.016135111451148987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,512,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,6144,0.011056000159846412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,4096,0.009920888476901585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,3584,0.008660444782839881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,5120,0.009696000152164036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,6144,0.015853333804342482
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,3072,0.00792711145348019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,4096,0.008336000144481659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,5120,0.013005333642164866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,3584,0.007997333175606197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,4096,0.012421333127551608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,3072,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,3584,0.011063111325105032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,2048,0.006158222340875202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,3072,0.009352888498041365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,2560,0.007366221812036302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,1536,0.004544888933499654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,2048,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,2560,0.007849777738253275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,1024,0.003432000055909157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,1536,0.006602666858169768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,768,0.0029395555870400537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,10240,0.020870221985710993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,1536,0.0069875551594628235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,1024,0.00628266649113761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,512,0.00254133323000537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,1024,0.005933333188295364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,768,0.005943111247486538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,6144,0.014467555615637036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,768,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,512,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,512,0.005450666778617435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,128,0.0018933332628673976
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,256,0.005614222337802251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,64,0.001782222133543756
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,256,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,128,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,32,0.001810666587617662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,128,0.004912888838185204
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,65536,0.0869564414024353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,2560,0.006864000111818314
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,64,0.005218666460778978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,32,0.005209777918126848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,51200,0.06764533122380574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,65536,0.05877244472503662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,16384,0.025943999489148457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,65536,0.05930133660634359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,12288,0.02087733315096961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,2048,0.00793511089351442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,51200,0.06555022133721246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,16384,0.019394665956497192
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,10240,0.01699555582470364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,16384,0.026072000463803608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,8192,0.015582222077581616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,12288,0.0159608887301551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,12288,0.02204622162712945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,7168,0.013817778064144982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,10240,0.01442399952146742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,256,0.0021822222818930945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,8192,0.012490666574902005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,10240,0.020072888996866014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,6144,0.012259555359681448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,5120,0.011095110740926532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,8192,0.0169840008020401
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,7168,0.011680000358157687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,4096,0.009659555223253038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,6144,0.01035288886891471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,7168,0.015271110667122735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,3584,0.01072266697883606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,6144,0.014718222949239941
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,4096,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,5120,0.012074666718641916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,3072,0.009372444616423713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,4096,0.011176000038782755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,3584,0.007712889048788283
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,2560,0.007999999655617608
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,3584,0.009712888962692684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,3072,0.007611555357774098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,2048,0.007286222444640265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,2560,0.007326222128338284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,3072,0.008645333349704742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,1536,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,2560,0.00793422261873881
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,2048,0.0069671107663048645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,51200,0.04725155565473768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,1024,0.004648000001907349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,768,0.004168888760937585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,1536,0.006640888750553131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,2048,0.007272000114123027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,512,0.0031884445084465873
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,1536,0.00693866651919153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,768,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,1024,0.005976888868543837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,256,0.0021840000732077495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,768,0.0058408888677755995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,512,0.005767999837795894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,512,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,256,0.005633777628342311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,64,0.0018044445249769424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,256,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,32,0.0018942221585247253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,128,0.005626666463083691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,5120,0.008824889030721452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,128,0.004879999905824661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,65536,0.07831911245981853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,64,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,51200,0.06161510944366455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,32,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,16384,0.023530667026837666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,65536,0.056792000929514565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,65536,0.05789688560697767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,51200,0.04627911249796549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,51200,0.05945510996712578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,16384,0.019239111079110038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,10240,0.015782222151756287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,16384,0.025248888466093276
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,8192,0.013432888521088494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,12288,0.021149333980348375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,10240,0.014144000079896716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,1024,0.006251555350091722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,7168,0.012846222354306115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,10240,0.018541332748201158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,6144,0.011567110816637674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,8192,0.011855111353927188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,8192,0.016529776983790927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,128,0.0019066666977273093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,7168,0.011177777416176267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,5120,0.009992000129487779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,7168,0.013409777647919126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,6144,0.010016000105275048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,4096,0.008632000121805403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,5120,0.008681777450773451
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,6144,0.013608888619475894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,3584,0.009461333354314169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,5120,0.01105955574247572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,4096,0.008127999802430471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,3072,0.008611555728647444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,3584,0.007753777835104201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,4096,0.010046222143703038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,2560,0.0074142225914531285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,3072,0.007679999702506595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,3584,0.008656000097592672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,2048,0.006313777632183499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,3072,0.00738933351304796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,2560,0.007267555428875818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,1536,0.005255110975768831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,2048,0.006960888703664144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,2560,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,1024,0.004566222015354368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,12288,0.015847999188635085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,2048,0.0069297779765393995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,1536,0.006631111105283101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,1536,0.006617777877383762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,1024,0.006275555739800136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,512,0.003140444349911478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,768,0.005952888892756567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,256,0.0021831111775504216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,1024,0.0059262220230367445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,768,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,512,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,128,0.0018275556051068837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,512,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,256,0.005600000007285013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,64,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,32,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,256,0.005208000126812193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,128,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,128,0.004861333303981357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,65536,0.0659502214855618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,51200,0.05212533473968506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,64,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,32,0.004918222212129169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,16384,0.019681778219011094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,65536,0.05676355626848009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,65536,0.05324000120162964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,12288,0.016160888804329764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,12288,0.018203554881943595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,51200,0.045886221859190196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,16384,0.018938667244381376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,10240,0.013770666387346057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,16384,0.023959111836221483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,768,0.0035075553589397003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,8192,0.011935110721323224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,12288,0.019724445210562814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,10240,0.013728000223636627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,7168,0.011036444041464063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,10240,0.017511111166742113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,8192,0.011739555332395764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,6144,0.010327111515733931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,7168,0.010679110884666443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,5120,0.009104000197516548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,6144,0.009367999931176504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,7168,0.01344622257683012
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,4096,0.00793599999613232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,6144,0.0124177775449223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,5120,0.008649778034951951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,3584,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,5120,0.01036622209681405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,4096,0.00903200027015474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,3072,0.008476444416575963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,3584,0.007979555262459649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,3584,0.008049777812427944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,2560,0.007190222541491191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,3072,0.007608888877762689
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,2048,0.006238222122192383
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,51200,0.0572213331858317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,3072,0.007335111498832703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,12288,0.015480889214409722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,1536,0.005485333502292633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,2560,0.0069253332912921906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,2048,0.006953777538405524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,1024,0.004247111164861255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,1536,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,768,0.003773333297835456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,1024,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,1536,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,512,0.003185777821474605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,8192,0.014094221923086377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,1024,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,768,0.0058906666106647914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,256,0.002491555487116178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,512,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,768,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,128,0.0021973333011070886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,512,0.0052888890107472735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,256,0.005559111220969095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,64,0.0017964444640609955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,4096,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,128,0.005572444448868434
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,32,0.0019048889064126545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,256,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,65536,0.05845599704318576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,128,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,51200,0.046182221836513944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,64,0.005056888692908817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,32,0.00508622204264005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,16384,0.017864000466134813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,65536,0.05594222413169014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,2560,0.0070426662762959795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,12288,0.014845333165592618
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,16384,0.018702222241295707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,51200,0.05004000001483493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,2048,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,10240,0.012227555943859948
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,12288,0.014919110470347934
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,8192,0.010661332971519895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,12288,0.0177182224061754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,10240,0.012842666771676807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,7168,0.009883555273214975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,10240,0.015828443898095023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,8192,0.010888000329335531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,6144,0.009362666971153682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,8192,0.013839110732078552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,7168,0.010348444183667501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,5120,0.008262222011884054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,7168,0.012100444071822695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,6144,0.009378666679064432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,4096,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,6144,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,5120,0.008629333641793992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,3584,0.006779555645253923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,5120,0.00868977771864997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,4096,0.008049777812427944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,65536,0.05156355433993869
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,3072,0.008336000144481659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,4096,0.008103110724025303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,51200,0.044907556639777295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,2560,0.0069546666410234236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,3072,0.007814221911960179
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,3584,0.007311111523045435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,2048,0.005944889038801193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,16384,0.02131555477778117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,3072,0.007165333463086023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,1536,0.0052426668504873914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,2560,0.006686222222116258
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,2048,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,1024,0.0041751112374994485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,1536,0.00675999994079272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,2048,0.006302222195598815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,768,0.0035013332962989807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,1536,0.006144000010357962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,1024,0.006276444428496891
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,512,0.003173333282272021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,1024,0.005620444400442972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,768,0.005962666538026597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,256,0.0025155554629034465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,512,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,768,0.0053582220441765254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,128,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,512,0.005211555709441503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,256,0.005602666487296422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,64,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,32,0.0018071112119489247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,256,0.005176888985766305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,128,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,65536,0.049103998475604586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,128,0.004937777916590373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,51200,0.03877422213554382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,64,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,16384,0.014820444915029736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,32,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,3584,0.007780444290902879
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,65536,0.05495199892255995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,12288,0.011677333050303988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,51200,0.04455822375085619
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,10240,0.010677333507272931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,16384,0.01832088828086853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,51200,0.04923289020856222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,12288,0.014436443646748861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,16384,0.020629333125220407
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,10240,0.012105777859687805
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,7168,0.008650666309727563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,10240,0.013712888790501488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,8192,0.012738666600651212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,6144,0.007975111405054728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,7168,0.010035555395815108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,7168,0.010397333237859938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,5120,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,6144,0.009472000102202097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,6144,0.00982844415638182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,4096,0.008640000389681922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,5120,0.008662222160233391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,5120,0.007981333467695449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,3584,0.007652444144090016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,65536,0.04961244596375359
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,4096,0.00796088907453749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,2560,0.0070675553547011475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,3072,0.005930666708283954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,3584,0.007961778177155389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,8192,0.009376889301670922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,2560,0.0052666668262746595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,12288,0.01535288823975457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,3072,0.007617777420414819
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,8192,0.010836444795131683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,2048,0.006254222244024277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,3072,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,2560,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,1536,0.005212444398138259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,2048,0.006976000136799282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,1024,0.003908444609906939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,1536,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,2048,0.006280888699822956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,768,0.0039013334446483185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,1536,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,1024,0.00630044440428416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,512,0.0031928889867332247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,1024,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,768,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,256,0.002480888946188821
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,768,0.005283555636803309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,4096,0.007995555798212687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,128,0.0021671110557185277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,512,0.004964444372389051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,256,0.005432888865470886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,64,0.0021991110924217436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,256,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,32,0.0018604445374674266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,3584,0.007319110963079665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,128,0.004944889081848992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,65536,0.035232888327704534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,64,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,51200,0.029275556405385334
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,32,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,65536,0.05436888999409146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,65536,0.04855022165510389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,2560,0.006628444625271692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,51200,0.0440613329410553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,12288,0.01036266651418474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,51200,0.04372088776694404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,16384,0.016570667425791424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,10240,0.00866844422287411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,16384,0.017837334010336135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,12288,0.013408888545301227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,8192,0.007717333734035492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,12288,0.01475822263293796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,10240,0.012079111403889127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,512,0.005625777774386936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,7168,0.00798044436507755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,10240,0.012409778104888068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,8192,0.010685332947307162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,8192,0.0105608892109659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,128,0.00563822231358952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,7168,0.009341333475377824
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,5120,0.00620977787507905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,6144,0.009370666411187913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,6144,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,4096,0.005960888746711943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,5120,0.00868444475862715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,3584,0.0055440002017551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,4096,0.008031999899281396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,5120,0.007977777885066139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,3072,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,16384,0.012049777640236748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,2560,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,4096,0.007400888535711501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,3584,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,2048,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,3072,0.007669332954618666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,3584,0.007299555672539606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,1536,0.0046862223082118565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,2560,0.007280000381999546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,3072,0.0069360000391801195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,1024,0.0035422220826148987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,2048,0.006981333096822103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,1536,0.006595555692911148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,2048,0.006128888991143968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,768,0.003208000005947219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,1024,0.006269333263238271
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,512,0.002883555574549569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,1024,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,768,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,6144,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,256,0.002193777718477779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,768,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,7168,0.010013333625263637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,128,0.002159111201763153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,512,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,64,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,256,0.005581333405441708
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,512,0.005161777966552311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,32,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,128,0.005575110928879843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,65536,0.025109334124459162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,256,0.004875555634498596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,128,0.004914666629499859
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,51200,0.02033955521053738
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,64,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,65536,0.054084443383746676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,32,0.004929777648713854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,16384,0.013096888860066732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,65536,0.04658044377962748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,12288,0.011016888750924004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,51200,0.04378844300905863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,51200,0.041197333070966936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,16384,0.015808888607554965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,10240,0.010026666853162978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,16384,0.01628444426589542
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,12288,0.013090666797426013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,8192,0.009019555317031013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,2560,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,7168,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,10240,0.011749332977665795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,10240,0.011043555206722684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,8192,0.01056977775361803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,6144,0.007305777735180325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,8192,0.009894222021102905
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,7168,0.009972444838947719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,5120,0.006255110932721033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,7168,0.009357333183288574
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,6144,0.009332444104883406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,6144,0.00869333330127928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,5120,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,3584,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,5120,0.00794933322403166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,4096,0.0079297779334916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,3072,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,3584,0.0075911109646161394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,3584,0.007107555866241455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,2560,0.004872000051869286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,3072,0.00758755538198683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,3072,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,2048,0.005935110979610019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,2560,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,2560,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,1536,0.005234666582610872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,12288,0.013249778085284762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,2048,0.005979555762476391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,2048,0.0069120000633928515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,1536,0.005764444255166584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,1536,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,768,0.0032248888164758682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,1536,0.005738666488064661
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,1024,0.006099555641412735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,4096,0.005648889061477449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,1024,0.0052639999323421055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,768,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,256,0.0024906665914588505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,768,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,512,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,128,0.002179555594921112
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,512,0.005223999834722943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,256,0.005365333209435145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,64,0.0018151110659042995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,256,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,32,0.00198399999903308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,128,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,65536,0.01943288909064399
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,128,0.004736888739797804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,51200,0.0162782222032547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,64,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,65536,0.05379466546906365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,32,0.00490399988161193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,16384,0.01276711126168569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,65536,0.045863999260796436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,51200,0.04282222191492716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,1024,0.00387999994887246
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,16384,0.015793777174419828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,51200,0.0387697782781389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,10240,0.008445333275530074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,16384,0.01479466590616438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,8192,0.006917333437336816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,12288,0.013088000317414602
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,512,0.0032008888406885993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,10240,0.011757333245542316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,7168,0.0069857777820693115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,8192,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,10240,0.011014222270912595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,6144,0.006280000011126201
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,7168,0.009690666364298927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,8192,0.009789333575301701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,5120,0.005593777944644292
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,4096,0.007361777954631382
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,7168,0.009296889106432596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,4096,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,5120,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,6144,0.008584889272848764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,3584,0.005584888988071018
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,4096,0.007939555578761632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,5120,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,3072,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,4096,0.007322666545708974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,2560,0.0052062223354975385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,3072,0.007623111208279927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,3584,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,2048,0.004193777839342753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,3072,0.0069324444565508105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,12288,0.009491555392742157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,2560,0.007287999822033777
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,2560,0.006592000110281839
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,2048,0.006958222223652734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,2048,0.006250666661394968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,12288,0.012070222033394707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,1536,0.006623999940024481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,1536,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,768,0.0031582222630580268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,1024,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,1024,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,6144,0.009292444421185387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,768,0.005990222096443176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,768,0.005285333428117964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,256,0.0021760000122918025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,512,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,512,0.004933333231343163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,128,0.0022035555707083806
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,256,0.005607999861240387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,256,0.004943111290534337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,64,0.0018364443547195857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,128,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,32,0.0018231111268202465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,1536,0.00452888881166776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,128,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,65536,0.027845333019892376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,64,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,51200,0.022169777088695105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,1024,0.0035031110876136353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,32,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,65536,0.04964888758129544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,16384,0.009338666995366415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,65536,0.044828444719314575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,12288,0.007594666547245449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,51200,0.03839733203252157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,16384,0.015842666228612263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,51200,0.03665688965055678
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,10240,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,16384,0.01442577772670322
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,768,512,0.002833777831660377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,12288,0.013057777451144325
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,12288,0.012088889049159156
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,10240,0.012078222301271228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,7168,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,8192,0.01035288886891471
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,10240,0.010996444357766045
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,6144,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,8192,0.009642666412724389
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,7168,0.009704888694816166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,5120,0.005280888742870755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,7168,0.009001777403884465
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,6144,0.009328888522254096
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,4096,0.005798222290145026
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,5120,0.008459555606047312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,6144,0.00867733359336853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,5120,0.008003555238246918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,4096,0.008011555506123437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,3072,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,4096,0.007271111011505127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,3584,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,3584,0.007252444823582967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,2560,0.004940444396601783
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,3072,0.007284444239404466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,2048,0.004547555413511065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,3072,0.006962666908899943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,1536,0.003559999995761447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,2560,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,2560,0.006578666468461354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,2048,0.006613333192136552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,1024,0.00309599998096625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,1536,0.0063688887490166565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,8192,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,2048,0.005690666536490123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,1536,0.005945777727497949
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,1024,0.006240889016124938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,512,0.0025111111915773815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,768,0.005623999983072281
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,1024,0.005210666606823604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,3584,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,768,0.0052897776994440295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,512,0.0057555557125144536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,128,0.0018479999982648427
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,512,0.005201777650250329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,256,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,64,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,32,0.0018088887963030075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,256,0.004734222259786394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,128,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,64,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,32,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,51200,0.013901333014170328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,65536,0.042118221521377563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,65536,0.046160889996422656
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,16384,0.006279110908508301
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,51200,0.03732088870472378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,51200,0.033596442805396184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,768,0.0026986667265494666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,16384,0.014193778236707052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,16384,0.015820443630218506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,256,0.0021964444054497615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,12288,0.01311555587583118
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,10240,0.005564444594913059
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,12288,0.012103111379676394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,8192,0.005304000029961268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,10240,0.011711111499203576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,128,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,7168,0.005274666680230035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,10240,0.010850666297806634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,65536,0.01679733395576477
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,6144,0.0047031111187405055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,8192,0.0103502223889033
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,8192,0.009700444837411245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,5120,0.004583999928500918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,7168,0.009715555442704095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,7168,0.009026666482289633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,6144,0.009204444785912832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,4096,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,5120,0.008341333104504479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,6144,0.008656000097592672
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,3584,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,5120,0.008027555214034187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,4096,0.00797333319981893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,3072,0.004560889055331548
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,4096,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,3584,0.007656888829337225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,2560,0.0045422220395671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,12288,0.004960000101062986
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,3072,0.007324444750944774
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,2048,0.004509333521127701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,3584,0.006959999601046245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,1536,0.0034799998005231223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,3072,0.0069413334131240845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,2560,0.0069191112286514705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,1024,0.0029306666304667792
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,2560,0.006498666687144174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,2048,0.0057760001056724125
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,768,0.002530666689078013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,1024,0.006046222315894232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,1536,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,512,3584,0.0052791109515561005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,1024,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,768,0.005921777751710679
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,256,0.002161777681774563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,768,0.004921777794758479
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,512,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,128,0.001855111163523462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,512,0.004892444445027245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,256,0.0052684446175893145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,256,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,128,0.005596444424655702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,32,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,65536,0.011043555206722684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,128,0.004548444516128964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,64,0.004552888787455029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,51200,0.009377777576446533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,65536,0.04556355542606778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,65536,0.041816890239715576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,16384,0.0053031109273433685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,51200,0.037031110790040754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,2048,0.0069146665434042616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,12288,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,51200,0.03328888946109348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,1536,0.006589333216349284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,10240,0.0052782222628593445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,16384,0.015564444992277356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,16384,0.014175999495718213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,512,0.002475555572244856
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,8192,0.0052373334765434265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,10240,0.011699555648697747
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,12288,0.01203555530971951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,7168,0.004880889008442561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,10240,0.011031111081441244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,8192,0.010400888820489248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,6144,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,8192,0.00963199966483646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,7168,0.009768889182143742
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,256,64,0.0018026667336622875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,7168,0.008991111483838823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,6144,0.009346666435400644
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,4096,0.00423911131090588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,6144,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,5120,0.008300444318188561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,3584,0.004616888860861461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,5120,0.007710222568776872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,3072,0.004523555437723796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,4096,0.007274666594134436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,32,0.004975999808973736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,2560,0.004896000027656555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,3584,0.00703022215101454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,3072,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,2048,0.004234666625658671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,3072,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,2560,0.00702311098575592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,1536,0.00349866681628757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,2560,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,2048,0.006624889042642381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,1024,0.0028293333533737394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,12288,0.012916444076432122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,2048,0.006304889089531369
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,1536,0.006610666712125142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,768,0.0025111111915773815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,1536,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,1024,0.005958222266700532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,512,0.0025111111915773815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,1024,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,768,0.005917333480384614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,256,0.002124444478087955
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,5120,0.004518222063779831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,768,0.005482666608360078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,128,0.0018355554590622585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,512,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,512,0.004600000050332812
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,64,0.0018275556051068837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,128,0.005230222311284807
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,4096,0.007667555577225155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,256,0.004864888886610667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,128,32,0.001822222231162919
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,65536,0.006244444598754247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,51200,0.005947555518812603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,3584,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,16384,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,64,0.004602666530344221
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,12288,0.004915555732117759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,32,0.004876444323195351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,10240,0.004535111288229625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,65536,0.04112622141838074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,51200,0.03355111016167535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,16384,0.014137778017255994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,7168,0.004605333424276776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,12288,0.011767999993430244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,6144,0.004608889006906086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,10240,0.0107360002067354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,5120,0.004217777815130022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,8192,0.009318222602208456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,4096,0.003958222352796131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,7168,0.008741333252853816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,3584,0.004559999952713649
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,6144,0.008369777765538957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,3072,0.0042444442709287005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,5120,0.00777599960565567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,4096,0.006979555719428592
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,2560,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,256,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,3584,0.0069679998689227635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,1536,0.0035582222044467926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,3072,0.006406222366624408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,128,0.004966222163703707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,1024,0.0026977778308921387
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,2560,0.006278222219811545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,768,0.002477333363559511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,2048,0.005928888916969299
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,512,0.0021751111166344746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,256,0.0018399999373488957
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,1536,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,128,0.0018719999740521114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,8192,0.004884444591071871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,768,0.005232000102599462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,64,0.0017804443422291016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,512,0.004920888692140579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,32,0.0018328889790508484
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,65536,0.006075555665625467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,128,0.004594666676388847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,51200,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,64,0.004608889006906086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,32,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,16384,0.005223111146026188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,12288,0.005046222358942032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,65536,0.04105777872933282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,10240,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,51200,0.03359466791152954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,8192,0.004224888980388641
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,16384,0.014099554883109199
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,64,2048,0.004210666649871402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,7168,0.003987555702527364
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,12288,0.011755555868148804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,6144,0.0041306668685542215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,5120,0.0038533334930737815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,4096,0.0038533334930737815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,8192,0.009374221993817223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,3584,0.0042257776690853965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,7168,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,6144,0.008009778128729926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,5120,0.007726222276687622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,2560,0.004543111142185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,4096,0.007071110937330458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,2048,0.003891555385457145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,3584,0.006986666884687211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,1536,0.0034959999223550162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,3072,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,1024,0.0028613333900769553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,2560,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,768,0.0024986666523747975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,2048,0.005705777969625261
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,512,0.0021457777669032416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,1536,0.005599110904667113
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,256,0.0021671110557185277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,1024,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,128,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,768,0.00534400012758043
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,512,0.004898666507667965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,64,0.0017804443422291016
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,256,0.004944889081848992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,32,0.001834666563404931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,10240,0.01070577816830741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,128,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,64,0.004527111020353106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,2,32,3072,0.0042160000238153665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,32,0.004592000196377437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,1024,0.005589333259397083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,256,0.004606222112973531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,16384,0.33092623286777073
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,12288,0.24982666969299316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,16384,0.1552248928281996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,51200,1.1154968473646376
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,10240,0.21207466390397814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,12288,0.12559999359978571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,8192,0.16977600256601968
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,10240,0.10559022426605225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,51200,0.45195643107096356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,7168,0.14152710967593723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,8192,0.08955111106236775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,6144,0.12221688694424099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,7168,0.08307200007968478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,6144,0.07220800055397882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,5120,0.10370222065183853
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,16384,0.17953422334459093
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,5120,0.06281510988871257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,12288,0.1903111139933268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,4096,0.08568444516923691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,51200,0.5196853213840061
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,4096,0.0515324440267351
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,3584,0.07276799943712023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,10240,0.13151644335852727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,3584,0.04589777853753832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,3584,0.058464891380733915
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,3072,0.062808891137441
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,8192,0.1178462240431044
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,2560,0.05251111255751716
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,7168,0.10206755664613511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,2048,0.04240355557865567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,3072,0.050887111160490245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,2560,0.03429066803720262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,2560,0.04389333393838671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,1536,0.03250844279925028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,2048,0.028956443071365356
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,2048,0.036257776949140764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,1024,0.0233271105421914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,1536,0.02849066588613722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,768,0.017851556340853374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,1024,0.017847999930381775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,4096,0.06793155272801717
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,512,0.012753778033786349
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,1024,0.020209777686331008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,768,0.015125332607163323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,256,0.007598222129874759
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,768,0.01664622293578254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,512,0.012828444441159567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,128,0.005079999979999331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,256,0.010421333213647207
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,256,0.009486222432719337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,64,0.004171555654870139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,128,0.010418666733635796
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,3072,0.040403554836908974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,32,0.004228444563017951
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,128,0.007962666451931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,64,0.007839110990365347
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,32,0.008010666403505538
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,6144,0.08714399735132854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,5120,0.0766577786869473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,65536,0.5516755315992568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,51200,0.42903823322719997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,16384,0.2502888838450114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,51200,0.8439786699083117
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,12288,0.18688533041212294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,16384,0.14504888322618273
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,1536,0.02365511159102122
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,16384,0.1292311085595025
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,65536,0.43887821833292645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,10240,0.15569955772823757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,12288,0.09733688831329346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,512,0.01276266657643848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,8192,0.1272302203708225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,12288,0.12176356050703262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,10240,0.0927271114455329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,8192,0.09198755688137478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,10240,0.12179556157853867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,8192,0.07136266761355929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,7168,0.11105244689517552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,6144,0.09611555602815415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,5120,0.08770755264494155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,7168,0.07958044608434041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,6144,0.05839911434385511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,6144,0.06918399863772921
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,7168,0.06810844606823392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,3584,0.06185866726769341
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,5120,0.06240355306201511
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,4096,0.041947556866539844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,4096,0.05550311009089152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,3072,0.05384799838066101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,3584,0.04656533400217692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,51200,0.3379528787400987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,3072,0.04281599985228645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,3072,0.03274844421280755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,2560,0.045420444673962064
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,2048,0.036031110419167414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,65536,1.0823627048068576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,2560,0.028431110911899145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,2560,0.03624711102909512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,1536,0.02816533380084568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,1536,0.019903999235894945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,2048,0.029913776450686987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,1024,0.020008888509538438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,1536,0.023338667220539514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,768,0.014780445231331719
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,5120,0.05075733198059929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,1024,0.015629332926538255
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,2048,0.02436977790461646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,512,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,1024,0.017166222135225933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,768,0.013401777380042605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,256,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,3584,0.03647733396954007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,768,0.014339556296666464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,128,0.003192000091075897
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,512,0.011023999916182624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,64,0.0024951110697454875
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,256,0.009369778136412302
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,512,0.01110311100880305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,32,0.002575111058023241
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,256,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,128,0.006931555353932911
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,64,0.0069368887278768755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,128,0.009376889301670922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,32,0.0069777775141927935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,65536,0.3158080048031277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,51200,0.2442613442738851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,65536,0.20302755302853057
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,16384,0.0833466649055481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,51200,0.16796089543236625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,12288,0.06346133020189074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,65536,0.17190399434831408
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,51200,0.1245359977086385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,16384,0.05076977941724989
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,16384,0.06901777452892728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,12288,0.04077422287729051
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,12288,0.05122133427196079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,8192,0.04352444410324097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,10240,0.036898665957980685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,7168,0.03842844565709432
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,10240,0.04127911064359877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,6144,0.031160000297758315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,8192,0.03053955568207635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,8192,0.03743288914362589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,7168,0.0299582216474745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,5120,0.027480888697836135
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,6144,0.026289777623282537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,7168,0.03309066759215461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,4096,0.022703111171722412
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,6144,0.028191109498341877
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,3584,0.019368888603316415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,5120,0.025056888659795124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,4096,0.019574221637513902
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,3072,0.01721333298418257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,4096,0.022239110536045496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,3584,0.01847288840346866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,3584,0.0199644449684355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,2560,0.015474667151769003
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,3072,0.01702044407526652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,4096,0.07170222202936809
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,2048,0.012329777909649743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,3072,0.017616889543003507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,2560,0.014918223023414612
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,1536,0.00997688869635264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,2560,0.015824000040690105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,2048,0.013344888885815939
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,1024,0.007350222104125553
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,768,0.006118222243256039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,2048,0.01384800010257297
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,1536,0.01143555591503779
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,1024,0.009027555584907532
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,512,0.0041697778635554844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,768,0.008261333737108443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,1024,0.009311111436949836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,256,0.0027555556346972785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,768,0.007968000239796108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,512,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,256,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,512,0.006271111054552927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,64,0.002201777779393726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,256,0.005574222240183089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,5120,0.022960889670583937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,128,0.006623111251327727
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,128,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,64,0.005304000029961268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,65536,0.23767466015285918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,10240,0.05407377746370104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,65536,0.1858675612343682
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,51200,0.18464355998569063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,1536,0.010859555668301053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,16384,0.06489955716662936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,65536,0.13395644558800593
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,51200,0.14299999343024358
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,12288,0.049330668316947095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,51200,0.11404800415039062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,16384,0.043749332427978516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,128,0.002185777864522404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,10240,0.041914665036731295
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,16384,0.05160533388455709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,12288,0.03571466604868571
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,12288,0.045760889848073326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,8192,0.03650844428274367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,32,0.002141333288616604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,10240,0.03179733289612664
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,7168,0.03223911258909438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,10240,0.03860444492763943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,6144,0.02818044357829624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,8192,0.026465778549512226
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,7168,0.02591733303334978
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,5120,0.02484888831774394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,8192,0.03228444523281521
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,32,0.005567111074924469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,4096,0.020958221620983545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,6144,0.023183110687467787
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,3584,0.0184560004207823
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,5120,0.019645333290100098
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,4096,0.017291555802027386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,5120,0.022210665875011023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,3072,0.016220443778567843
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,3584,0.01584533353646596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,4096,0.02013066742155287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,2560,0.01391377713945177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,3584,0.017842666970358956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,3072,0.014496000276671516
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,2048,0.012038222617573209
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,3072,0.016415999995337594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,2560,0.013106666505336761
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,1536,0.008096888661384583
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,2048,0.011534222298198275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,1024,0.00637511122557852
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,2048,0.012503999802801343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,1536,0.009709333380063375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,768,0.00526311124364535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,1536,0.010777778095669217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,1024,0.008289777570300633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,512,0.0031644445326593188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,1024,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,768,0.0074275558193524676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,256,0.0024497778051429326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,768,0.006976889239417181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,512,0.007229333122571309
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,128,0.0021555556191338436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,7168,0.02831733226776123
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,512,0.006161777923504512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,256,0.0064222220745351576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,64,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,256,0.005254222287072076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,32,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,128,0.006257777826653586
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,128,0.00508088908261723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,64,0.0052915554907586836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,32,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,51200,0.15666133827633327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,65536,0.15553155210283068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,2560,0.014472888575659858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,16384,0.054901334974500865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,65536,0.12277955479092068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,51200,0.12522133191426596
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,12288,0.04214488797717624
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,10240,0.03611822260750665
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,16384,0.04930488930808174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,51200,0.10085778103934394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,12288,0.031714667876561485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,8192,0.030265778303146362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,12288,0.037267555793126426
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,7168,0.027477333943049114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,10240,0.030702223380406696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,8192,0.02495999965402815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,6144,0.024562666813532513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,8192,0.027239999837345544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,6144,0.024753777517212763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,5120,0.021370665894614324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,7168,0.02312533391846551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,7168,0.023211555348502264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,6144,0.020451555649439495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,4096,0.018072888255119324
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,65536,0.20125155978732637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,6144,0.020272889071040683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,5120,0.017998221847746108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,4096,0.015554666519165039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,3072,0.015108444624476962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,4096,0.01688799924320645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,3584,0.01494488947921329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,2560,0.013618666264745923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,3584,0.015202666322390238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,3072,0.013419555293189155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,2048,0.009882666998439366
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,3072,0.013793778088357715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,16384,0.0391937792301178
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,2560,0.011807999677128263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,1536,0.008621333373917473
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,2560,0.012084444363911947
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,2048,0.010055999788973067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,1024,0.0057911111248864075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,2048,0.011043555206722684
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,10240,0.028024001253975764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,1536,0.009026666482289633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,768,0.003997333347797394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,1024,0.007653333246707916
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,512,0.0033893332713180115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,768,0.007291555404663086
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,1024,0.007643555601437886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,256,0.002393777792652448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,768,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,512,0.006691555596060223
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,5120,0.018392889036072623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,128,0.0021475555582178966
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,3584,0.017031111650996737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,512,0.005894222193294101
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,256,0.00629066675901413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,32,0.0018515555808941524
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,256,0.0052879999081293745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,128,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,128,0.004912000149488449
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,64,0.005024000174469418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,65536,0.16400888231065539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,32,0.005072000126043956
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,51200,0.1256951093673706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,16384,0.044166223870383374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,65536,0.11597243944803874
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,65536,0.07152622275882296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,51200,0.08851910961998834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,51200,0.05737777551015218
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,16384,0.02428177826934391
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,10240,0.029535998900731403
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,16384,0.03478844298256768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,1536,0.009143110778596666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,12288,0.02027555637889438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,8192,0.022616000639067754
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,12288,0.027951998843087092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,7168,0.020961778031455144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,10240,0.018760888112915885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,8192,0.01615111033121745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,10240,0.026611554953787062
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,6144,0.019208000765906442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,7168,0.015826667348543804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,64,0.0019306666735145782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,5120,0.01516888870133294
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,7168,0.018573333819707234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,6144,0.014431110686726041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,4096,0.013136000268989138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,6144,0.016716443830066256
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,5120,0.013273778061072031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,3584,0.011992000043392181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,4096,0.011680000358157687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,5120,0.01444355481200748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,3072,0.01038044442733129
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,4096,0.013170666992664337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,3584,0.010939555863539377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,2560,0.009600888523790572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,3584,0.012438221938080259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,3072,0.009682666924264696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,2048,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,12288,0.03460622164938185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,2560,0.008967111508051554
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,3072,0.011039999624093374
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,2560,0.010401777923107147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,2048,0.007988444632954067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,1536,0.007612444460391998
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,2048,0.009278222090668148
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,768,0.003528888854715559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,1536,0.007601777712504069
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,512,0.002883555574549569
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,1024,0.006776000062624614
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,1024,0.006572444405820634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,768,0.006588444527652528
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,256,0.0024542222834295696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,8192,0.020388444264729817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,768,0.005948444621430502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,256,0.0058871110280354815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,256,0.005179555465777715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,128,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,128,0.004888888862397936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,32,0.0018088887963030075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,64,0.00490488898422983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,32,0.004935111022657818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,1536,0.006688888702127669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,65536,0.14425599575042725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,1024,0.004543999830881755
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,51200,0.1135751141442193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,65536,0.0653066635131836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,16384,0.04029066695107354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,51200,0.08628088898128933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,128,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,51200,0.05326666765742832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,12288,0.03231911195649041
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,512,0.005561777700980504
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,10240,0.026707554856936138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,16384,0.03274666600757175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,64,0.001871111078394784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,12288,0.026608000199000042
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,8192,0.02279377811484867
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,10240,0.015589333242840238
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,10240,0.02272266646226247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,7168,0.01795733306143019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,8192,0.01406311160988278
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,8192,0.0203795548942354
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,6144,0.015817777978049386
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,7168,0.012886222038004132
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,7168,0.01811911165714264
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,5120,0.013635555903116861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,6144,0.012079999678664737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,65536,0.1011235581503974
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,6144,0.01588800052801768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,4096,0.011687999798191918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,5120,0.011047110789351992
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,5120,0.013749333719412485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,3584,0.010440000229411654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,4096,0.010039110978444418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,3072,0.009333333207501305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,4096,0.012768000364303589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,16384,0.021902221772405837
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,3584,0.009205333060688442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,2560,0.008284444610277811
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,12288,0.017774222625626456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,3584,0.011789333489206104
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,3072,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,2560,0.007948444121413762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,3072,0.011061333119869232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,1536,0.00563466673096021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,2048,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,1024,0.003957333250178231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,2048,0.008348444269763099
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,1536,0.006759111252095964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,768,0.0035208890007601846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,1536,0.0073520003093613525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,1024,0.006305777778228124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,512,0.0028408887899584244
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,1024,0.006236444330877728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,768,0.006253333141406377
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,256,0.0021946666141351065
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,512,0.005950222412745158
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,768,0.005920889063013925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,128,0.0020248889923095703
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,512,0.0052595556610160405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,256,0.005587555468082428
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,512,0.006241777704821692
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,64,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,256,0.004936000125275718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,128,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,128,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,65536,0.12210755878024632
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,64,0.004911111046870549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,32,0.004882666799757215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,51200,0.0950853361023797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,2560,0.010055999788973067
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,16384,0.036277333895365395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,65536,0.06332000096638997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,12288,0.028311110205120508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,51200,0.05099022057321337
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,51200,0.07441155778037177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,16384,0.02090222140153249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,10240,0.023969777756267126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,16384,0.030966222286224365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,8192,0.01993333299954732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,12288,0.017116443978415597
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,12288,0.024000888069470722
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,7168,0.01818933255142636
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,10240,0.021377777059872944
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,8192,0.01350400017367469
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,6144,0.016306666864289176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,8192,0.0188773340649075
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,7168,0.012512889173295764
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,32,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,7168,0.01648888819747501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,6144,0.011753777662913004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,4096,0.010819555156760745
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,6144,0.014911111858155994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,5120,0.010416000253624385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,2048,0.007308444215191736
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,3584,0.009923555784755284
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,5120,0.013328889177905189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,4096,0.008989333278603023
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,4096,0.012392889294359418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,65536,0.09220533238516913
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,2560,0.007899555895063613
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,3584,0.011443555355072021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,3072,0.008032889001899296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,3072,0.010491555763615502
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,2048,0.006601777755551868
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,2560,0.007595555649863348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,1536,0.00555377784702513
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,2560,0.009331555830107795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,2048,0.007063999772071838
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,10240,0.01577333278126187
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,1024,0.00397511116332478
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,2048,0.007983110845088959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,768,0.003548444559176763
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,1536,0.0069146665434042616
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,1536,0.006980444822046492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,768,0.006075555665625467
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,1024,0.006228444476922353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,5120,0.014198222094111972
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,512,0.0028231110837724474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,256,0.002198222196764416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,768,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,512,0.006046222315894232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,128,0.0019680000841617584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,256,0.005721777677536011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,512,0.0053280000057485365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,3072,0.00865955568022198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,256,0.005236444373925527
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,64,0.001830222292078866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,128,0.005582222094138463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,128,0.004537777768241035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,65536,0.10496888558069865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,64,0.004859555512666702
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,32,0.004874666531880696
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,51200,0.08189866940180461
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,65536,0.08264355527030097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,65536,0.06064977910783556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,16384,0.031239999665154353
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,51200,0.04911377694871691
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,51200,0.06897599829567803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,16384,0.02016888890001509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,1024,0.006258666515350342
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,10240,0.020977778567208186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,16384,0.026683555708991155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,8192,0.018872888551818002
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,12288,0.016511999898486666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,10240,0.01493866741657257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,12288,0.022944000032213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,7168,0.016329778565300834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,8192,0.013110222087966071
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,10240,0.019894222418467205
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,6144,0.014683556225564746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,8192,0.017610667480362784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,7168,0.012395555774370829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,5120,0.012878222597969903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,6144,0.01093066649304496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,4096,0.009869332942697737
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,3584,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,5120,0.009677333136399588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,6144,0.014368888404634265
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,3584,0.008594666918118795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,5120,0.01274577776590983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,4096,0.008440888590282863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,3072,0.007735110819339752
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,3584,0.008038221961922115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,4096,0.011762667033407422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,2560,0.007010666860474481
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,3584,0.010693333215183683
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,3072,0.007616000043021307
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,2048,0.005922666854328579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,3072,0.009707555174827576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,12288,0.02498400045765771
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,1536,0.004836444639497333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,2048,0.006940444310506185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,2560,0.008110221889283922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,1024,0.003626666549179289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,2048,0.007958222594526079
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,1536,0.006616000086069107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,768,0.003335111050142182
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,1024,0.006239999913507038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,1536,0.006820444431569841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,512,0.0025519999778932994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,1024,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,768,0.006232888748248418
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,256,0.0021848889688650766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,512,0.00589155571328269
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,768,0.005624889085690181
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,128,0.0020942222326993942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,256,0.005623111294375525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,512,0.005290666802061929
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,64,0.001839111041691568
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,32,0.0017928888814316855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,128,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,128,0.004607111215591431
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,65536,0.08866222037209405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,64,0.004883555488453971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,51200,0.06678311030069987
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,32,0.004575110971927643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,16384,0.025736000802781846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,65536,0.07770222425460815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,65536,0.05858844518661499
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,12288,0.019080888893869188
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,2560,0.007307555940416124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,16384,0.01943199998802609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,51200,0.06223999791675144
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,10240,0.01789866718981001
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,12288,0.01605244477589925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,8192,0.015156444576051501
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,12288,0.021360889077186584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,10240,0.014469333820872836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,7168,0.013752889302041797
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,10240,0.019109333554903667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,8192,0.0127697777416971
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,6144,0.011742221812407175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,8192,0.016799110505315993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,7168,0.011471110913488599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,7168,0.016137777103318107
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,7168,0.015433778365453085
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,5120,0.010501333408885531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,256,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,4096,0.00927466650803884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,6144,0.010330666270520952
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,6144,0.013172444370057849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,3584,0.012068444656001197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,5120,0.008981333010726506
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,4096,0.008303999900817871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,3072,0.009711999860074785
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,4096,0.011328000161382886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,3584,0.007985778152942657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,2560,0.009008000294367472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,3584,0.009707555174827576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,51200,0.04713866776890225
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,2048,0.007803555164072249
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,3072,0.007674666742483775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,3072,0.008923555413881937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,16384,0.026829333768950567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,1536,0.006288888967699475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,2048,0.006937777830494775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,1024,0.004535111288229625
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,1536,0.006593777901596493
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,1536,0.006295111030340195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,768,0.0041831110914548235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,1024,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,1024,0.005613333235184352
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,512,0.0032115555885765287
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,768,0.005904888941182031
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,768,0.005250666704442766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,256,0.002171555534005165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,512,0.005903999838564131
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,512,0.005004444470008214
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,128,0.001871111078394784
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,256,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,256,0.004905777672926585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,64,0.0018293333964215384
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,128,0.005583111196756363
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,32,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,128,0.004543111142185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,2560,0.007139555282062954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,65536,0.0776968863275316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,2560,0.007635555333561367
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,64,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,51200,0.061016890737745494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,16384,0.02362844513522254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,2048,0.007317333585686154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,65536,0.05759021970960829
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,65536,0.07062577539020114
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,12288,0.018227555685573153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,51200,0.04665866825315687
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,16384,0.018985778093338013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,51200,0.05471200082037184
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,16384,0.02473955518669552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,12288,0.01583377851380242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,8192,0.013414222333166333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,12288,0.02091911103990343
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,10240,0.01424888935354021
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,7168,0.01293955577744378
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,10240,0.01979200045267741
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,8192,0.01238933371173011
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,6144,0.010803555448849997
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,7168,0.011035555766688453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,5120,0.010478222535716163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,7168,0.013341333303186627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,6144,0.00977955593003167
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,4096,0.008581333690219456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,6144,0.0125937776433097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,5120,0.008980444735950893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,3584,0.011032000184059143
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,5120,0.01181688904762268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,4096,0.008400888906584846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,3072,0.008970666262838576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,4096,0.010354667074150508
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,3584,0.008031999899281396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,32,0.004586666822433472
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,2560,0.00906044410334693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,3584,0.008328888979223039
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,3072,0.007607111500369177
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,3072,0.008000888758235509
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,2560,0.007347555624114142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,1536,0.006248000181383557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,2560,0.007556444240940942
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,2048,0.00721866637468338
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,2048,0.007259555160999298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,1024,0.004463110946946674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,1536,0.006590222318967183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,768,0.003752888904677497
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,1536,0.0063386667105886675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,1024,0.006252444452709622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,512,0.00315022220214208
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,5120,0.01219911128282547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,1024,0.005913777897755305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,768,0.005991999887757831
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,256,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,512,0.00590133335855272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,128,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,256,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,512,0.004946666873163647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,256,0.004909333255555895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,64,0.001828444500764211
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,128,0.005551110953092575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,32,0.001810666587617662
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,128,0.004606222112973531
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,65536,0.0668026672469245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,2048,0.00741155528359943
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,64,0.004942222187916438
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,10240,0.015792888071801927
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,32,0.004583111239804162
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,51200,0.05292977889378866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,16384,0.020276443825827706
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,65536,0.05673422416051229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,65536,0.06978844271765815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,12288,0.01620533400111728
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,51200,0.055600888199276395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,16384,0.01891911029815674
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,10240,0.014142221874660917
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,16384,0.022172444396548804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,8192,0.015503999259736804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,8192,0.012096000214417776
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,12288,0.015463110473420886
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,12288,0.019310222731696237
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,10240,0.013806222213639153
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,768,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,7168,0.011137777732478248
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,8192,0.011395555403497485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,10240,0.016000888413853116
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,8192,0.015080889066060385
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,7168,0.01069155583779017
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,5120,0.009301332963837517
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,6144,0.010010666317409938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,7168,0.01256444470749961
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,4096,0.008057778080304464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,6144,0.012044444680213928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,5120,0.009024000002278222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,3584,0.007351111206743453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,4096,0.008277333445019191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,3072,0.009205333060688442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,4096,0.008314666648705801
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,3584,0.00794222205877304
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,2560,0.0075146663520071245
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,3584,0.008360000120268928
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,3072,0.007638221813572778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,2048,0.007277333074145847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,51200,0.04589244392183092
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,3072,0.007435555259386699
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,1536,0.0058959999846087555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,2048,0.0070257774657673305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,2048,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,1024,0.004599110947714912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,1536,0.00664533343580034
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,1536,0.005939555664857228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,6144,0.010205333431561789
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,768,0.0038951109680864546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,1024,0.006296000132958095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,768,0.005953777581453323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,512,0.003504000190231535
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,1024,0.005592000153329637
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,256,0.002540444541308615
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,768,0.0052355556852287715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,512,0.00591822216908137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,128,0.0022106667359670005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,5120,0.010681777364677854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,512,0.004939555707905027
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,64,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,256,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,32,0.00180177783800496
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,128,0.005615111026499007
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,65536,0.057384888331095375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,128,0.004572444491916233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,51200,0.04591200086805555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,2560,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,32,0.004878222114510006
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,2560,0.0069573331210348345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,16384,0.018245332770877414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,65536,0.055777774916754834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,65536,0.06031644344329834
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,12288,0.014447111222479077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,51200,0.045244445403416954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,10240,0.012228444218635559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,16384,0.018710222509172227
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,51200,0.05067111055056254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,12288,0.015056888262430826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,8192,0.01130222198035982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,10240,0.013148444394270578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,12288,0.016184000505341422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,7168,0.010384889112578498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,8192,0.010778666370444827
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,10240,0.01514844430817498
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,6144,0.009314667019579146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,8192,0.013472889032628802
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,7168,0.010048888623714447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,5120,0.008280889027648503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,7168,0.011500444677140979
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,4096,0.00759733302725686
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,256,0.005598222215970357
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,6144,0.01036977767944336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,5120,0.008692444198661381
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,3584,0.0069511110583941145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,5120,0.008438222110271454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,4096,0.008247999681366814
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,3072,0.007794666621420119
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,4096,0.007269333634111617
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,3584,0.008223999705579545
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,64,0.004895110925038655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,3584,0.007655111451943715
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,2560,0.007241778075695038
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,2048,0.005711110929648082
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,3072,0.007647111184067196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,3072,0.0069715554515520734
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,2560,0.007354666789372762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,1536,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,2048,0.006923555499977536
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,2560,0.006621333460013072
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,1024,0.004572444491916233
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,2048,0.00628355559375551
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,16384,0.021195555726687115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,1536,0.006625777731339137
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,768,0.003585777762863371
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,512,0.0031724443866146933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,1536,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,1024,0.006237333433495627
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,256,0.002534222271707323
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,768,0.005952000204059813
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,1024,0.005562666803598404
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,512,0.005664000080691443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,512,0.004948444250557157
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,256,0.005621333503060871
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,6144,0.009594666461149851
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,64,0.0018862223045693503
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,256,0.004563555535342959
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,128,0.005590222362014983
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,128,0.004571555389298333
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,65536,0.04784533381462097
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,64,0.004592000196377437
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,32,0.004581333448489507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,51200,0.03895644346872965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,65536,0.055139554871453174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,65536,0.05858577622307671
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,16384,0.014175999495718213
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,51200,0.04477599925465054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,12288,0.012131555212868584
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,51200,0.0477128889825609
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,16384,0.018256889449225534
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,10240,0.010471999645233154
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,16384,0.020609777834680345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,12288,0.014581332604090372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,8192,0.009020444419648912
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,12288,0.016715556383132935
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,10240,0.012068444656001197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,7168,0.008674666285514832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,10240,0.01332355539004008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,768,0.00519822206762102
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,6144,0.00794488853878445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,8192,0.010805333654085795
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,8192,0.012091555529170565
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,7168,0.010023111270533668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,32,0.001847111102607515
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,7168,0.01076533314254549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,6144,0.009376889301670922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,4096,0.00925333301226298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,6144,0.009573332965373993
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,5120,0.00867377801073922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,3584,0.008323555191357931
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,4096,0.00832177781396442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,5120,0.008086221913496653
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,3072,0.006544888847404056
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,3584,0.007627555893527136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,2560,0.005594666633341048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,3584,0.0070773329999711775
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,3072,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,2048,0.00628977765639623
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,3072,0.006596444381607904
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,2560,0.007332444190979004
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,2560,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,1536,0.00600622221827507
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,2048,0.0069448889957533936
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,1024,0.003867555409669876
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,2048,0.006102222121424145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,1536,0.006634666687912411
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,768,0.0038951109680864546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,1536,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,128,0.002181333386235767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,1024,0.006102222121424145
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,512,0.0028880000528362063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,1024,0.005241777747869492
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,256,0.002504000026318762
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,768,0.005604444278611078
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,5120,0.006994666324721442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,128,0.0021626665774318907
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,512,0.005617777920431561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,256,0.005569777968857024
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,64,0.001850666685236825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,512,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,128,0.00526577772365676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,256,0.004951111144489712
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,65536,0.036329779360029436
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,128,0.004556444370084339
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,4096,0.007613333563009898
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,32,0.004584889031118817
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,65536,0.05472266674041748
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,16384,0.012042666474978128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,65536,0.052207999759250216
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,51200,0.044179555442598134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,12288,0.010386666489972008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,51200,0.04266400138537089
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,16384,0.017647999856207106
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,10240,0.008691555923885768
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,12288,0.013419555293189155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,12288,0.014713777436150445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,8192,0.007636444436179266
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,10240,0.012053333222866058
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,10240,0.012117333710193634
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,7168,0.007807999849319458
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,768,0.005111111121045219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,8192,0.010785777535703448
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,6144,0.007198221981525421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,8192,0.01037688884470198
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,7168,0.010028444230556488
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,5120,0.006188444379303191
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,7168,0.009563555320103964
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,6144,0.009402666654851701
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,4096,0.006285333385070165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,51200,0.028577778074476454
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,6144,0.008714666797055138
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,64,0.00488711107108328
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,3584,0.005554666535721884
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,5120,0.008711111214425828
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,3072,0.005646222167544895
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,5120,0.007983999947706858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,4096,0.008329778081840938
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,2560,0.005249777601824866
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,3584,0.007941332956155142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,4096,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,16384,0.016511110795868766
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,2048,0.005633777628342311
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,3072,0.007672000262472365
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,3584,0.006990222467316522
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,1536,0.004591111093759537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,3072,0.006705777926577463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,1024,0.004210666649871402
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,2048,0.006956444846259222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,2560,0.006232000059551663
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,1536,0.00665244460105896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,2048,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,512,0.0028657778683635923
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,1536,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,1024,0.0062871111763848206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,256,0.0021884443445338146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,768,0.005896889087226655
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,1024,0.005277333160241445
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,128,0.0021688888470331826
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,512,0.005723555468850666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,768,0.0052826665341854095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,64,0.0018853334089120228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,512,0.00489777781897121
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,256,0.00555644432703654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,32,0.0018435555199782054
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,256,0.004635555462704765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,65536,0.024893333514531452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,128,0.004613333278232151
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,51200,0.020266667008399963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,64,0.004573333180612988
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,65536,0.05425777700212267
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,16384,0.014022222823566861
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,65536,0.05092444353633457
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,51200,0.04377422067854139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,12288,0.011399110986126794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,51200,0.03733955489264594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,16384,0.015835555063353646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,10240,0.010072000324726105
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,2560,0.007295110987292395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,12288,0.013108444710572561
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,8192,0.009287111461162567
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,12288,0.012655110822783576
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,10240,0.011772444678677453
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,7168,0.008227555288208855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,10240,0.011410666836632622
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,8192,0.010444444086816577
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,6144,0.007344000041484833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,7168,0.010022222167915767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,7168,0.008997333546479544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,5120,0.006606222026877933
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,128,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,6144,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,4096,0.00592533333433999
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,6144,0.008440888590282863
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,5120,0.00868000007337994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,3584,0.00555288874440723
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,32,0.004566222015354368
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,4096,0.00793511089351442
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,5120,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,3072,0.005577777822812398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,4096,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,2560,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,3584,0.008007999923494127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,3072,0.007312888900438945
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,3584,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,3072,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,2048,0.006626666833957036
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,2560,0.007304000357786815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,1536,0.005560889012283749
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,2560,0.006068444500366847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,768,0.0035128887328836652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,2048,0.0069271110826068455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,1024,0.004230222354332606
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,2048,0.00591466658645206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,1536,0.006631999793979857
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,768,0.0038355555799272326
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,1536,0.00527022240890397
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,1024,0.006215999937719769
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,512,0.003160888950030009
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,8192,0.009677333136399588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,1024,0.0052328887912962176
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,768,0.005570666657553778
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,256,0.00227111101978355
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,128,0.0021644443687465456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,768,0.005024888863166173
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,256,0.005221333354711533
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,64,0.0018560000591807896
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,128,0.005431111074156231
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,256,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,32,0.0019057778020699818
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,128,0.004631999880075455
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,64,0.004558222161398994
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,32,0.004885333279768626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,51200,0.01648622254530589
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,65536,0.04583555459976196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,16384,0.015767110718621146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,65536,0.05384977658589681
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,16384,0.014671999547216626
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,51200,0.04347999890645345
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,12288,0.009733333355850643
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,16384,0.015835555063353646
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,51200,0.0369102226363288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,10240,0.00851644410027398
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,16384,0.015277332729763456
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,12288,0.013045333325862885
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,8192,0.007253333098358578
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,10240,0.011747555600272285
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,12288,0.012078222301271228
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,7168,0.007290666302045186
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,512,0.005573333137565189
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,10240,0.01090400003724628
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,6144,0.006581333362393909
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,512,0.004870222260554631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,8192,0.010451555252075195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,5120,0.0052151112920708126
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,7168,0.009714666340086196
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,8192,0.009372444616423713
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,4096,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,6144,0.009367110828558603
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,65536,0.01960000064637926
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,7168,0.009014222357008193
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,3584,0.005548444473081165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,5120,0.008617777791288164
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,3072,0.005247999810510212
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,6144,0.008652444514963362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,4096,0.008234666453467475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,5120,0.00792622235086229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,2560,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,2048,0.0041831110914548235
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,3584,0.00795911086930169
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,4096,0.006924444602595435
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,3072,0.007599111232492659
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,1536,0.005139555368158552
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,3584,0.006978666616810693
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,2560,0.006973333656787872
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,1024,0.0038737778862317405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,3072,0.006635555376609166
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,2048,0.0069431112044387394
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,768,0.0034560000316964257
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,1536,0.006222222414281633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,2560,0.005924444645643234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,512,0.002834666727317704
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,2048,0.0059004442559348206
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,1024,0.006091555373536215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,1536,0.005552000055710475
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,768,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,1024,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,512,0.005470222069157495
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,768,0.0052168890833854675
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,256,0.005506666584147348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,64,0.0018337778747081757
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,32,0.001867555495765474
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,128,0.005244444641802046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,256,0.00462400002612008
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,65536,0.027828445037206013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,128,0.004577777865860197
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,51200,0.02267644471592373
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,64,0.004594666676388847
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,65536,0.05200711223814222
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,16384,0.009119999905427298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,65536,0.04518577787611219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,51200,0.03836888737148709
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,12288,0.007937777373525832
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,51200,0.036375999450683594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,16384,0.015489778584904142
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,16384,0.014132445057233175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,10240,0.006959111326270633
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,12288,0.012921777864297232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,12288,0.011763555308183035
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,8192,0.0052382221652401825
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,10240,0.011417778001891242
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,256,0.0021955555097924336
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,10240,0.010743111371994019
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,7168,0.005568888866239124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,8192,0.01036711119943195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,6144,0.0059368887709246735
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,512,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,7168,0.010009778042634329
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,5120,0.0052391112678580815
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,7168,0.008913777768611908
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,6144,0.009311111436949836
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,4096,0.005906666732496685
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,6144,0.008368000388145447
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,5120,0.008325333396593729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,32,0.004867555366622077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,3584,0.005252444495757421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,5120,0.007989333735571967
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,4096,0.008031999899281396
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,3072,0.004923555586073133
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,4096,0.006950221955776215
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,2560,0.004919111314747068
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,3072,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,3584,0.006569777925809224
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,2048,0.004574222283230888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,3072,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,2560,0.007263110743628607
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,1536,0.003858666867017746
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,2560,0.006048000107208888
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,2048,0.006716444260544247
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,1024,0.0029102222373088202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,2048,0.005600888695981767
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,1536,0.006577777779764599
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,768,0.0028159999185138275
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,768,128,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,1536,0.005585777676767773
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,1024,0.0058755555914507965
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,8192,0.009660444325870937
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,512,0.0024871110088295406
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,256,0.002191999927163124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,1024,0.00508266646001074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,768,0.005601777798599667
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,512,0.005606222069925732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,128,0.0018533333722088074
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,64,0.0018248889181349014
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,768,0.00489688871635331
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,256,0.005281777845488654
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,512,0.004690666579537922
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,512,32,0.0018444444156355327
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,65536,0.016163556112183463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,128,0.005559999909665849
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,256,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,51200,0.014006222287813822
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,128,0.0042053332759274375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,16384,0.005956444475385878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,64,0.00425155543618732
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,65536,0.046543111403783165
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,32,0.004590222405062782
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,3584,0.00793688909875022
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,12288,0.00490133340160052
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,51200,0.03705688979890611
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,65536,0.04186755418777466
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,10240,0.0052586665583981415
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,51200,0.03424888849258423
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,8192,0.004926222066084544
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,16384,0.015495111544926962
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,7168,0.004924444274769889
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,12288,0.01313866674900055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,16384,0.013970666461520724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,10240,0.01167555567291048
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,6144,0.004588444613748127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,12288,0.011743110915025076
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,8192,0.010390222072601318
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,10240,0.010680889089902243
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,5120,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,4096,0.0042462220622433555
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,7168,0.010040000081062317
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,8192,0.009340444372759925
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,3584,0.004917333523432414
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,7168,0.008671999805503422
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,5120,0.008647999829716152
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,3072,0.004515555583768421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,4096,0.00832088871134652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,6144,0.00832266691658232
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,5120,0.0079644446571668
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,2560,0.004881777697139316
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,3584,0.007654222349325816
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,4096,0.007213333414660559
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,2048,0.0046160001721647055
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,3072,0.007648000286685095
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,3584,0.006567111031876669
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,1536,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,2560,0.006974221931563483
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,1024,0.0028284444577164124
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,3072,0.006614222294754452
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,2048,0.006630222416586346
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,2560,0.0063582224150498705
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,1536,0.006607999818192587
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,2048,0.00565244464410676
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,1536,0.0052577778697013855
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,1024,0.005772444523043103
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,512,0.002505777817633417
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,1024,0.004930666751331753
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,256,0.002144888871245914
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,512,0.005537777725193236
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,512,0.004585777719815572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,128,0.0018880000958840053
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,256,0.005227555417352253
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,256,0.004576888763242298
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,64,0.0017911110901170308
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,128,0.005256888767083486
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,32,0.0018044445249769424
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,6144,0.009314667019579146
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,128,0.004231111043029361
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,64,0.004545777622196409
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,32,0.004506666627195147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,51200,0.009567110902733272
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,65536,0.041125333971447416
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,256,768,0.0025333333760499954
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,16384,0.005576000031497743
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,51200,0.033246222469541765
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,768,0.005937777873542573
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,12288,0.00489155575633049
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,768,0.005271111097600725
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,16384,0.014211555322011312
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,10240,0.004916444420814514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,12288,0.013094222380055321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,12288,0.011755555868148804
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,8192,0.0052488889131281115
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,10240,0.011886222494973077
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,10240,0.01074044406414032
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,7168,0.004886222382386525
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,8192,0.010438222024175854
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,8192,0.009531555904282464
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,6144,0.004903111192915174
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,7168,0.010030222435792288
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,7168,0.009150221943855286
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,65536,0.045871112081739635
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,5120,0.004261333495378494
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,6144,0.009454222189055549
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,6144,0.008380444513426887
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,4096,0.004583999928500918
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,51200,0.037032888995276556
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,3584,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,5120,0.007622222105662028
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,4096,0.008301333420806462
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,3072,0.004609777695602841
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,16384,0.015465777781274585
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,4096,0.007316444483068254
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,3584,0.007663110891977946
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,2560,0.004576000074545543
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,3584,0.0062773335311147906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,2048,0.00425866660144594
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,3072,0.00720266666677263
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,2560,0.00703288863102595
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,3072,0.006648889018429651
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,1536,0.0035208890007601846
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,2048,0.006504888749784893
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,2560,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,65536,0.010797333386209277
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,768,0.0025253333151340485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,1024,0.005946666830115848
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,1536,0.0052560000783867305
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,512,0.0022960000981887183
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,1024,0.004877333425813251
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,768,0.00591555568906996
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,256,0.0021280000607172647
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,512,0.005538666827811136
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,768,0.004589333302444882
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,128,0.0018453333112928604
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,512,0.004899555610285865
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,5120,0.008672888908121321
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,256,0.0052675555149714155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,64,0.001793777777089013
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,32,0.001842666624320878
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,256,0.004531555705600315
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,128,0.00554044461912579
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,65536,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,128,0.0042444442709287005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,64,0.004342222379313575
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,16384,0.004855999930037392
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,32,0.004879111217127906
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,12288,0.004591111093759537
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,51200,0.03305244445800781
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,10240,0.00424177779091729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,128,1024,0.0028693332440323303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,8192,0.004908444566859139
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,16384,0.013759999639458127
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,2048,0.0053119998839166425
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,7168,0.004852444347408083
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,6144,0.0039013334446483185
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,10240,0.010337777435779572
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,5120,0.0041626666982968645
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,8192,0.009324444664849175
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,7168,0.008652444514963362
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,6144,0.007982222570313348
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,5120,0.007334222396214803
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,3584,0.0042444442709287005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,4096,0.006584888945023219
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,3072,0.003875555677546395
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,51200,0.006246222390068903
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,2560,0.00453244439429707
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,3584,0.0063564446237352155
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,2048,0.004197333421972063
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,65536,0.04105777872933282
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,3072,0.006157333238257303
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,2560,0.005902222047249476
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,1024,0.00282577777074443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,1536,0.00665511108107037
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,768,0.002566222308410539
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,12288,0.011720889144473605
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,1536,0.005222222043408289
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,512,0.002493333278430833
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,256,0.002161777681774563
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,1024,0.004938666605287128
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,128,0.0018577778504954444
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,4096,0.003925333420435588
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,768,0.005246222019195557
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,512,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,32,0.0017706666969590718
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,256,0.00489066665371259
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,65536,0.006311999840868845
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,128,0.004263111286693149
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,51200,0.0059279998143513995
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,64,0.004910222358173794
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,16384,0.005261333452330695
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,32,0.00423555572827657
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,12288,0.004506666627195147
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,65536,0.04140799906518724
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,10240,0.004515555583768421
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,51200,0.03290844294759963
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,8192,0.0042053332759274375
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,1536,0.0035128887328836652
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,7168,0.004226666771703296
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,12288,0.011406222151385413
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,6144,0.00423200014564726
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,5120,0.0041751112374994485
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,10240,0.010704889065689512
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,4096,0.004190222256713443
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,2048,0.005558222118351195
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,8192,0.009095999929640029
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,3584,0.004200889004601372
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,3072,0.003903999924659729
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,7168,0.008651555412345463
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,6144,0.00794311116139094
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,2560,0.004242666479614046
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,5120,0.007630222373538547
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,4096,0.00699733363257514
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,1536,0.003516444315512975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,3584,0.006576889091067844
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,64,64,0.0015902222237653202
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,1024,0.002809777855873108
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,768,0.0024986666523747975
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,2560,0.005659555395444234
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,512,0.00217333332531982
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,2048,0.005607111172543631
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,256,0.0021386666016446217
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,1536,0.005605333381228977
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,1024,0.004986666556861666
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,128,0.0018417777286635505
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,64,0.0017946666727463405
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,768,0.00490044429898262
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,32,0.0016373333831628163
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,512,0.004559111264016894
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,256,0.004564444637960858
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,128,0.004567111117972268
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,16384,0.014152889450391134
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,64,0.00461422238085005
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,32,0.004465777840879229
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,float16,1,32,2048,0.0038951109680864546
VLLM,0.19.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,3072,0.005948444621430502
